SECOND

SECOND
Writing scientific papers and giving talks at meetings and conferences
are essential parts of research scientists' work, and this short, straight-
forwardly written book will help workers in all scientific disciplines to
present their results effectively. The first chapter is about writing a
scientific paper and is a revision of an essay that won first prize in a
competition organized by Koch-Light some years ago. Later chapters
discuss the preparation of manuscripts, speaking at meetings and
writing theses. One chapter is for scientists whose first language is not
English . Another is addressed to those in North America . The last
chapter gives information about dictionaries, style books and other
literature.
COMMUN I CA TING IN SCIENCE
Writing a scientific paper and
speaking at scientific meetings
COMMUNICATING IN
SCIENCE
Writing a scientific paper and
speaking at scientific meetings

Second edition
VERNON BOOTH
Formerly of Trinity Colltgt, Cambridge

CAMBRIDGE
UNIVERSITY PRESS
CAMBRIDGE UNIVERSITY PRESS
Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, Slo Paulo

Cambridge University Press

The Edinburgh Building, Cambridge CB2 2RU, UK

Published in the United States of America by Cambridge University Press, New York

www.cambridge.org
Information on this title: www.cambridge.org/9780521429153

<0 Cambridge University Press 1984, 1993

This publication is in copyright. Subject to statutory exception

and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without
the written permission of Cambridge University Press.

First published 1984

Second edition 1993
Tenth printing 2006

A catalogue record/or this publication is availablefrom the British Library

Library of Congress Cataloguing In Publication data

Booth, Vernon .
Communicating in science: writing a scientific paper and speaking at scientific meetings I
Vernon Booth. - 2nd ed.
p. cm.
Includes bibliographical references and index.
ISBN 0-52 1-42915- 3 (pbk.)
I. Communication in science. 2. Lectures and lecturing.
3. Technical writing. I. Title.
Q223.B664 1993
808'.0065-<lc20 92- 37481 CIP

ISBN-13 978-0-521-42915-3 paperback

ISBN- I 0 0-521-42915-3 paperback

Transferred to digital printing 2006

Dedicated to T . W. Fline
Contents

.
page XI.. Foreword
XII BalkJons & instructicns for the typist & the printer
XIII Explanad<m for the seCQnd editWri
xv Gkkssary ofsome printers' tenns
1 Chapter One: Writing a scientific paper
28 Chapter Two: Before you lecture or talk to us, please read this
41 Chapter Ex: Empty numbers
44 Chapter F.our: Preparation of the script and figures
50 Chapter Five: Addressed to those for whom English is a
foreign language
54 Chapter Six: An appeal to North Americans
64 Chapter Seven: Preparation of a doctoral dissertation or thesis
69 Chapter Eight: Further reading
75 Index

.
lX
Foreword

This final edition of 'Communicating' was assembled by Cambridge

University Press from material prepared by Vernon Booth before his
death in 1991. H is family would like to thank the Press for thus making
publication possible.

.
XI
 ..
XII

Balloons & instructions for

the typist & the printer
Don't
type
fMH

ta it

( Ok
a l~
Explanation for the second edition

In one laboratory in Cambridge, if a person became unapproachable,

we said he o r she was giving birth. Paper labour can be a traumatic
experience, but should not be. The writing of a paper, or a book,
although indeed a task, should be a pleasant occupation. Books on
scientific writing have been published, but scientists 'do not have time'
to read them. So, in 1970, I wrote an essay, Writing a Scien tific Paper,
and submitted it for a competition organized by Koch- Light Ltd; the
essay was awarded first prize, and issued as a booklet.
Later editions of the booklet grew longer and were published by the
Biochemical Society. For the CUP version , various sections were
expanded into chapters. This made the book longer, but the principal
chapter re mai.ned short and kept its original title. The subtitle of the
previous CUP edition was writing and speaking. This was felt to be
misleading; so it has been changed. Numerous other changes have been
made for this edition.
Chapters One and Two are intended primarily to help scientists,
engineers and others to write papers for journals and to give sho rt
talks. However, nearly all the suggestions also apply to the writing of
books and the delivery of full lectures.
The style, especially of the first chapter, is succinct , at times even
terse. So much had to be written, in so small a space, that conciseness
was highly desirable. Chapte r One is not suitable for fast reading.
Parts of the book are written in the imperative, the simplest style.
This is not intended to be categorical. True, certain parts are con-
troversial; but life would be dull if we all agreed. There may be errors;
most books have errors. But I ask you to read it (as an examinee once
added) E. & o.E. (errors and omissions excepted).

Some of the words that are discussed are in 'quotes' or italic. A

plethora of quotes and italic can be irritating; so, where the meaning
XIII
 .
XIV EXPLANATION FOR THE SECOND EDITION

should be clear without them, they are not used, even at the risk of
some loss of consistency.
Examples of a directive being discussed are referred to in brackets.
Thus (1 (10)) means the re is an illustrative example or more infor-
mation in the line marked (10) in the margin of Chapter One .
You will see T . W . Fline mentioned in various places. This refers to
Those Whose First Language Is Not English. Whenever we write or
speak , we must think of these people.
The majority of papers submitted for publication are returned to
authors for revision. Naturally, you would like each of your papers to
be accepted without change. This book cannot guarantee yo ur fulfilling
that ambition, but perhaps it will help.
As you read this book, you may realize that I enjoyed writing it. I
offer best wishes that you too will enjoy writing, preparing scripts and
speaking.

I am grateful to many, many colleagues, as well as to several editors

at CUP and in various countries, for help and suggestions.

Vernon Booth
January 1991
Glossary of some printers' terms

balloon. Ring drawn round instructions to the printer. [Page 46.)

bold. Heavy type as here. In a script, underline with a ~a.~x!~~~·
braces. Curly brackets { } .
brackets. Square brackets [ ). The term bracket is used in a general way
to include parentheses ( ), b races { } and angle brackets ( ) . To an
author brackets usually mean parentheses.
caps, upper case.CAPITAL LETTERS . In a script underline
three times.
copy. The script. To avoid confusion, a photocopy should be so named.
em rule. Long dash(-) . Length of cap M. (1 (33).) Many publishers
use a spaced en rule ( - ).
en rule. Short dash (-). Half the length of em rule. [Page 21. J
foliation . Numbering of folios.
folio. (1) Sheet of script. (2) Page number . (3) Sheet of paper of any
size folded once.
full point. Full stop, period.
index. (1) Alphabetical list of topics at the end of a book. Plural,
indexes. (2) See superior below.
inferior, subscript. Small low digit(s) or other character(s). H 2 S04 , 2n·
ital, italic. Sloping type. Spelt with lower-case 'i' . In a script, underline
once.
justified lines. Lines of print made the same length by varying the
spaces between words. See window below .
leading. Space between lines of type. Pronounced 'ledding' .
I.e., lower case. Small letters, i.e . not caps.
legend, caption. Explanation to a figure . Term occasionally also used
for explanation to a table. Ideally, legends should be understandable
without reference to the text, and, to identify them they are often set in
smaller type. Legends (or captions) to figures are usually placed below,

xv
xvi GLOSSARY OF SOME PRINTERS' TERMS

whereas those for tables are usually placed , more logically, above the
display.
letter space. Space be tween letters.
numeral. Digit. See page 16 under Homonyms.
par. Paragraph. (1 ( 16).] N.P. New paragraph. (Page xii.] Indicate NP
by 0 or ___s-.
parens. Parentheses, round brackets ( ). (1 (18).)
quotes are often called 'inverted commas' though only the first is that;
the second is an apostrophe or raised comma; 'single', " double"
reference marks. • t t § 11 1 •• t t . .. . Use them in this order for
footnotes.
reprint, offprint. A printed copy of a single article from a journal or
book. If available before publication then known as a preprint.
rom, roman. Normal upright type, not italic or bold. Spelt with
lower-case 'r'.
run on. Continue in same para. See last sentence, page xii.
sanserif, sans. Type without serifs. THIS is sans. H girder; 0 ring; S
shape; T join; U tube; V groove. For text, sans is less legible than
type with serifs. (8 (1).] See The typewriter's or word processor's
type face (page 47).
sm. cap, small caps. Capital-style letters only slightly larger than I.e.
Used for EMPHASIS , for HEADINGS and for some CONVliNTIONS. In a
script line, underline twice.
superior, superscript. Small high digit(s) or other ch aracter(s). mm 3 , in.
Also called index; plural , indices.
widow or club line. Short line at the top o f a page.
.
window. Wide, ugly gap between words an a line.
word space. Space between words.
 Chapter One
Writing a scientific paper

Before you write

Here are four suggestions about what you might do before writing a
paper.
I. If notebooks are used, good notebook discipline is helpful. When
an experiment is finished, try to record your conclusion in words,
together with your findings .and on the same page. Make tables. Draw
graphs and stick them into the book. Keep a file in which to record
summaries of results from many experiments, and group them by
subject. Some experiments will each provide results for various summa-
ries. Number each book and each right-hand page. Then , even after
some years, an experiment can be found from the file as e.g. 9;43
(book 9; p . 43). Write the date at the top of the page.
Prompt recording of a summary for each experiment compels you to
give critical thought to the experiment at the best time, and may move
you to repeat a control test while you still have the materials. Clark
(1.960) makes an eloquent appeal for keeping adeq uate notes. Write
every digit unmistakably. Think: 'I must so write my notes that another
person can read them if I am ill, or worse' . Then you should
understand them yourself when you come to write the paper.
2. Speaking makes you think out arguments; and listeners' criticisms
( I) may prevent your publishing a clanger. Some institutes operate a
regular tea club or occasional seminar at which researchers tell
colleagues about their work. If your institute has no club, or the
programme is filled, invite colleagues to your room to listen to you.
Display diagrams. If you have no projector, use a felt-tip pen to draw
diagrams and tables on the back of a roll of wallpaper. Hang the paper
over a chair on the bench. Do - speak - slowly.
(2) Nothing clarifies ideas in one's mind so much as explaining them to
other people.
I
 :Z. C OMMUNI CAT ING I N SCIENCE

3. The third suggested pre-writing activity is based on Woodford's

(8 (3)) 'reservoirs'. Take 8 sheets of paper. Boldly label them
Title Summary Intro Mat Meth Results Disc Ref
Write ideas for your paper, whenever they come to you, as notes on
the appropriate sheets (reservoirs). Use differently coloured sheets if
possible. Carry a card everywhere - even to bed. Jot down ideas as
they occur. Transfer the notes to the reservoirs and put a fresh card in
your pocket or handbag. Rewrite a cluttered reservoir from time to
time; if you wait too long, you may forget what some of your notes
meant. Hold the reservoirs in a clothes peg (pin), not in a wire clip
which may catch on other papers.
Some writers construct a skeleton , an outline scheme, before they
start to write. Should you do this it is still advisable first to prepare the
reservoirs. In particular, a skeleton for the Discussion may help you to
muster your ideas in the best order and to avoid repetition.
4 . Prepare tables and figures.

Honesty & accident

If the .result of an experiment seems 'wrong', record it none the less,
and watch for a repeat. Many a discovery has been made by accident:
serendipity is alive and productive. H owever, people have been known
to manipulate or 'doctor' their 'wrong' results. Manipulators may have
regrets later.

When to begin writing

My research supervisor said 'Writing a paper is as important as
experiments. Is it unreasonable , then, if it takes as long?' Oft-repeated
(3) advice is 'Set aside your paper for some weeks, then read it. You may
be amazed at what you wrote. ' You m ay even discover a passage you
yourself cannot understand. If you follow this advice, and believe that
supervisor, you must start writing early. Writing as the work proceeds
reveals gaps in knowledge, gaps that should be filled while laboratory
facilities are still available.

Arrangement of a scientific paper

The commonest arrangement for a research paper is that indicated by
the order of the reservoirs mentioned above. Some investigations are
suitable for results and discussion to be written together in narrative
form. If you use this form, write your Conclusion as a separate section.
 WRITING A SCIENTI FI C PAPER 3

Where to start
Even though you have enough material, you may have postponed
writing a projected paper. Perhaps you find it difficult to start. I do.
You do not have to begin with the Introduction. Begin with the easiest
section. This may be Methods, for you should know what you did. Use
the 'reservoirs', and cross out the notes as you consume them.
Next, perhaps, you might start on the Results. Write the first draft
'in your own words' j ust as though you are telling a friend about your
discoveries. Don't worry - yet - about grammar, aptest words & style.
The immediate objective is to get going. You can polish the style later.
This paragraph was so written , and the needless words and hackneyed
phrases have not yet been polished out.
The Conclusion of a paper is so important that you should make its
first draft in time to allow for re-draftings.

Stocktaking
Now take stock. The outline is complete, diagrams and tables are
ready, the Discussion is planned , the Conclusion is drafted and
Methods are written. Oh joy! the paper is half finished. A h appy
author writes better than a worried one.

Title & key words

Some searchers may read only a paper's Title and Summary. So both
are supremely important parts. Compose trial versions of the title as
early in your writing as you can; re-examine them later.
On your first reservoir sheet write key words for the Title . Let the
Title's first word be a key word if possible; in lists of titles such a word
is better than 'The'. Remove other waste words such as 'on', 'study',
'investigation' .... The Title should be short yet specific, no t general:
a reader , attracted by a title, may be disappointed to find that the
paper is about only one specialized aspect of the subject promised.
Have you experienced such a disappointment?
Many journals require a Headline or Running title as well as the
Title. An ingenious paraphrase of the Title can supplement the latter.
For example, the Latin name of a species might appear in the Title and
the common name in the Running title.
If the journal needs key words, you can pro vide them from your
.
reservoir.
 4 COMMUNICATING IN SCIENCE

Summary
If the editor permits, compose the Summary in numbered paragraphs.
The first should state - briefly- what you did. Then come the main
results. Lists of values may be indigestible for your readers; so use
words, supplemented by a few key values. State your conclusion in the
last paragraph. If you have no succinct conclusion, you might write
'The effect of A upon B is discussed' .
If a summary is long, readers may look only at the first and last para-
(4) graphs. Although a well-written summary may be lifted by abstractors,
a long summary will be shortened, perhaps by the omission of what you
consider vital parts.
Write the Summary in the past tense, except perhaps the last
paragraph.
Some journals print the Summary in small type . How odd!

Introduction to a paper
The Introduction should state the problem, and perhaps ask a ques-
tion. The objective must be clear. If you modified your objective after
you began the work , give the current version. Do you still think you
asked the right question?
The quoting of numerous papers in the Introduction is no longer
good practice. [If much has been published, and you think it warrants a
review, write that separately and submit it to an editor.] Refer to
papers that, taken together, show that a problem exists. If another
paper gives many references, refer to that. However, beware of lifting
references - from that paper - together with misquotations of infor-
mation from the original papers. That has been done .... For
example, one abstractor supposed that Kaninchen meant little dog; and
Yamane's work on the rabbit [Kaninchen means rabbit) has gone into
the literature as being on the dog. For this and other cautionary tales,
see Hartree (1976) . Roland {1976) reports that J. Hlava, a Czech,
wrote an article 'On dysentery' and added a Czech translation of the
title: 'O. Uplavici'. An abstractor cited the author as 0 . Uplavici; so an
author who never lived went into the literature for 50 years.
In the last sentence of the Introduction, it is accepted practice to
state the conclusion. A reader can better appreciate the evidence that
follows if it is clear what conclusion is being supported . However, this
version of the conclusion must be brief. Some authors repeat much of
the Summary in the Introduction. That is not an acceptable practice.
 WRITING A SCIENTIFIC PAPER S

Materials & Methods

If the description of materials is short.it may be included in Methods.
A void trade names if practicable, not to avoid advertising, b.u t because
they may not be understood abroad. [Do you know what Klampits or
Barbistors are? or what Skellysolve means?) If you use a local name for
polymethylmethacrylate or other compound, give the chemical name
at first mention of the trade name.
Write what you did in operational order. Invert 'The urn was dated
after restructuring' to 'The urn was reconstructed, then dated'. You
should so describe the methods you used that others can repeat the
experiments. You must be concise, yet you must not omit essential
detail. If you used 'alcohol' say which alcohol. If you controlled, or
even measured, the humidity and ventilation in an animal room, say so:
they may be nearly as important as temperature. If you centrifuged a
suspension, say whether supernatant or pellet was used for the next
operation. Similarly, if you filtered a suspension, say which part was
retained; it has not always been easy for a reader to conjecture.
If you used control experiments, permit no doubt about their nature.
The reader may not be able to guess what you omitted for each control.
If your paper is about a new method, ask a visitor or a technician to
test your description by applying the method in your absence. The
result of the omission of one detail can be illuminating.

Results
Before you write about your Results it may be advisable to study Units
and quantities (p. 25) and Tables (p. 24).
Replicate observations should not usually be given. Instead, offer the
mean and a measure of the variability if you can. The range is not
satisfactory; if there are enough replicates for the range to be of use
then there are enough for estimating the standard deviation (s.o.) of
one observation, the standard error of the mean (s.E.M.) or the
coefficient of variation (c. v.). Give the number of observations or the
(5) degrees of freedom within parentheses: 12.65 + 0.22 (n = 12). Perhaps
you can make a pooled estimate of the variance (or other statistic) from
the whole study. You can then give individual uncluttered values.
Journals ask for tables and figures to be clear without reference to
the text. This requires concise explanation in legends, an explanation
of abbreviations, and care in the avoidance of repetition in the text and
in other legends, as well as consistency between text and legends.
6 COMMUNICATING IN SCIENCE

Discussion
The Discussion must not be so long as to deter a reader, yet it must
contain logical argument. Do not repeat descriptions of other people's
findings if they are in the Introduction; refer to that. Avoid summariz-
ing your results in the Discussion. Mention them, take them as read or
refer to a table or even to the Summary (quote a paragraph number, if
used). Enlarge upon the significance of your new results and explain
how they add to existing knowledge. You may have formulated your
problem as a question in the Introduction. If you can now give the
answer, that facilitates discussion.
Think critically, not only about other people's work but about your
own. For example, ask yourself 'Can my hypothesis be refuted? Can
my results have another explanation?' Maier (1933) told the students in
one of two large groups that, were they unable to solve the problem
given to them, they should try to ignore their first approach and seek
an altogether different line. (The other group, the control, was not
told.) This worked - in the 'told' group a larger proportion solved the
problem than in the control group - yet it is difficult to achieve such
'lateral thinking', as de Bono (1967) calls the modern development.
The following example shows how important is such 'no-prejudice
rethinking' . Two authors published graphs to prove their thesis that
xanthine oxidase and the Schardinger enzyme (aldehyde oxidase) are
distinct enzymes. Later, their graphs were used by another author to
confirm the opposite (now accepted) view that the enzymes are
identical. Had those first authors given their results more thought, they
too might have reversed their conclusion. The literature contains
abundant examples of inconclusive thinking. Writers should take care
not to add to them by publishing in haste.
W. Pauli wrote 'I don't mind your thinking slowly: I mind your
publishing faster than you can think. ' (Translated by Mackay ( 1977). J

Conclusion
If you are fortunate, your Message (or part of it) may survive in
textbooks - although you may not be given a whole sentence! So the
Conclusion needs precise wording. Your Conclusion may appear
three times: in the Discussion, the Summary and the Introduction . Do
not repeat the wording; paraphrase it. If the reader has not understood
one version, another may help. Use the shortest version for the
Summary.
 WRITING A SCIENTIFIC PAPER 7

Parting remarks. Perhaps you have not yet reached a conclusion, but
have contributed towards one. You may like to end with 'Parting
remarks'. Make them short, but do not bring ' final' into the heading.
One paper concluded with 'We admit we have raised more questions
than we have solved.'

Reference$, Bibliography or Literature cited

Write each reference on a card. Arrange the cards in order and give
them to the typist at the final typing of your paper when you have
checked them against your revised text. This scheme leads to fewer
errors than does retyping the references at every retyping of the paper,
although the advent of word processors has made 'editing' the Refer-
ences a much easier job. Each journal arranges references in a
particular style, which should be followed. Give the typist a copy of the
journal to provide an example of style. If references in the text are
made by superior digits, avoid 'the value was 242 '; change to 'the value
was 24 (ref. 2)'.
Check the typed list against the original papers. Also check that the
spelling of names in text and Bibliography agree. Errors are very
(6) common in papers submitted to editors. (See Numbering, p. 48.)

Written English
Good written English is nearly the same as good spoken English.
Grandiloquence has no place in scientific writing. We need to convey
ideas effectively, to make it easy for the reader to understand what we
write, not to exhibit our vocabulary. Indeed, those who use pompous
language may even be suspected of having nothing important to say!
Try to envisage the reader; write in a manner not too technical , not too
elementary. Write as though you are talking to a reader, relating your
experiments, but restrain colloquialisms.

Clear English. Ask yourself often: would T . W. Fline [p. xiv)

understand what I write.? Write short sentences, but not all of them so
short as to produce a staccato effect. Cure a staccato passage by linking
two sentences (as I have done here with a 'but'), but do this only
seldom , so as to keep to 'one idea per sentence' with occasional
exceptions. A satisfying sentence has two main verbs (Perttunen, 1975).
If you train yourself to speak well, that will help you to develop a
good written style. In conversation, choose words with care, speak
deliberately and clearly.
8 COMMUNICATING IN SCIENCE

Grammar
English grammar is simpler than that of many languages, yet some
writers are careless about the small amount that does exist . People
abroad who learn English as a foreign language mostly learn it
grammatically because they are used to complex grammar in their own
languages. So when they meet doubtful grammar in a published work,
or at a conference, they may be confused. On their behalf, I appeal to
you to persuade emergent authors to follow rules of grammar and to
punctuate carefully.
In English, the same word can sometimes be used as a noun (tin), an
adjective (tin can) and a verb (to tin the copper wire). Indeed, almost
any noun may be 'verbed' and any adjective ' nouned': medical, high ,
sabbatical. But this freedom is needlessly abused. Consider the phrase
'the book was authored' . Why not say 'written'? 'Authored' gives no
special shade of meaning. Where a verb is needed but none exists, it is
practical to use a noun (to program, to chromatograph); but 'to gift'
(mineral samples were gifted by Dr Fob) is unnecessary.

Mutual editing
In courses on rapid reading, one is told not to go back to re-read a
passage. A trained reader may not return to a sentence whose meaning
was not grasped. How can you discover such passages in your own
writing? One way is to put the paper away for a month , then read it
afresh. This may be impractical. Another is to have colleagues read
your paper. Ask them both to make general comments and to mark
every sentence they had to read twice . If they are critical , thank them
nevertheless, for , if they fail to understand you, others might fail too
and your Message will be lost .
For nearly 2000 years it has been known that we see other people's
faults more easily than our own. (Parable of the mote and beam ,
Matthew 7 , iii.) Moreover, it is fuh to cross out needless words in other
people's papers. Therefore , make a deal with your colleagues; if they
will let you 'correct' their papers, you will let them correct yours. Do
not use red ink, which is offensive to some; green is more soothing.
You may have noted a repetition above. That is deliberate, because
emphasis is needed. Hundreds of the errors I have seen , in papers that
had already been accepted but not yet edited, ought to have been seen
by a critical colleague and then corrected - before submission.
However senior you be, ignore the 'statuskline', or hierarchy, and
 WRITING A SCIBNTIFIC PAPER 9
ask for constructive criticism, not flattery , from your juniors. Such
editing should be good training for them.

Literary style
(7) Noun adjectives. In English , nouns may be used as adjectives; that is,
as modifiers of a true noun. One might write 'an oil engine needs
engine oil' o r 'glass bottles are made of bottle glass'. But the use of
nouns as adjectives may lead to confusion unless made with care. For
such terms as 'dog meat' or 'cat fish' make it clear which of the two
meanings is intended. 'Rapid gas apparatus deterioration' is better
written as 'Rapid deterioration of gas apparatus', and 'product
treatment' as 'treatment of the product'. Even the simple 'drug dose'
seems clumsy. If you dislike recurrent 'of', the occasional genitive case
maybe used.
There is no suggestion here that nouns should never be used
adjectivally. Many are so used satisfactorily, including hydrogen bond,
gold size, oak tree, steel plate, SI units.
Take care to avoid confusion . We may know that a spring washer is a
sprung washer , not a machine for washing springs, but would T . W .
Hine know that? And would the title 'Cancer in rubber workers' be
understood?
When two or more nouns are used as adjectives of one noun , the
phrase may become inelegant. Consider the following: isotope dilution
assay results; pH 6.8 phosphate buffer; multiple conductor galvanized
angle steel pylon system; we devised a new short chain fluorocarbon
aerosol can valve. Such phrases are difficult to comprehend; the reader
finds that each successive noun is not the real noun; words have to be
stored mentally until the substantive being modified is reached. There-
fore, avoid long adjectival phrases, or stacked modifiers as Woodford
(8 (3)) calls them (see also p . 54). Even if the modifier contains no
noun adjective, it may be troublesome, as this example shows: a
frequently heated and therefore deeply coloured viscous solution . . ..
The use of hyphens may lead to improvement of some phrases, but
rewriting is usually best.
Note that in vivo, excess and de novo are not adjectives, but that
subliminal, optimal, minimal and maximal are. Write 'test in vitro' not
in vitro test. People would not write ' in glass test' .

Comparatives. A passage that contains a comparative SO!Jletimes causes

difficulty: what does 'lions eat more than antelopes' mean? Make clear
IO COMMUNICATING IN SCIENCE

what is more than what, and only compare things that are comparable.
Instead of 'starch yielded more glucose than maltose' write either• ...
than did maltose' or 'starch produced a greater yield of glucose than of
maltose'. Do not omit 'those in' from 'bearings in steam engines lasted
longer than those in diesel engines'.

Wrongly attached participle. A verb in the active voice needs a subject,

an operator, either actual or understood. One of the most common
errors submitted for publication is exemplified by ' having completed
the observations the telescope ... 'or 'a bend was observed in the
bridge using a strain gauge' . Was the bridge really using a gauge?
Phrases such as the following make people laugh. After standing in
boiling water for 2 h , examine the flask ; electronic devices should be
made safe before leaving the laboratory; and goggles are required to do
the experiment. Yet such errors (aberrations, faults, lapses ... ) are
often submitted to editors. Read what Fowler (8 (9)] or the authors of
style books listed in Chapter Eight have written on Unattached,
Wrongly attached or Dangling participles and infinitives.

Gerund. A participle may become a kind of noun (called a gerund), as

in 'Writing a paper'. If the adding of ' the' and 'of' (e.g. before and
after 'adding' in this sentence) makes grammatical sense, the -ing word
is a gerund. Applying this test, you can !'ee that 'Using a dyna-
mometer, the tractive effort was measured' is not allowable because
'Using' is a wrongly attached verb here, not a gerund , being wrongly
attached to 'effort'. Change the sentence to 'Using a dynamometer, we
measured . . .• or to 'A dynamometer was used for measuring the
tractive .. .'or add 'By' -+'by using a .. .'. In •Applying this test , you
... '(see 6 lines above) the 'you' was the subject, and the phrase is
allowable.

Be alert for dangling verbs. So many sentences start with 'Judging by'
or ' Based on' that these participles are becoming accepted for use in
that way.
Even so, let other words that end in -ing or -ed warn you to ensure that
each is either a gerund or is properly attached to an operator.
Using is written and spoken so often without an expressed operator
that it may eventually become a preposition (compare providing). Even
if you accept that event, you will be wise to ensure that the operator is
always clearly understood. The phrase 'He could not stop the mill using
 WRITIN G A SCIENTIFIC PAPER II

the brake' should not be written because it has two meanings. Add ' by'
before ' using'.
Schoenfeld (8 (7)) writes typically amusing pieces on using.

Waffte, needless words

(8) If you put aside your draft, and then examine it later, this is the time to
(9) expunge what Houp & Pearsall (8 (4)) call 'empty words'. Such phrases
as 'It is worth pointing out in this context that' may be deleted without
affecting the sense. So may the following.
It is significant to note the fact that
relevant to mention here that resu lts
reported here demonstrate that
It is known that
It should be borne in mind in this connection that
found to be
Such phrases, which correspond to spoken anderms, are not needed in
reports. Many phrases may be shorte ned. For 'It is plainly demon-
strable from the curves presented in Fig. 2' write 'Fig. 2 shows'. If a
piece is introduced by 'Needless to say', why say it? ' Recent' is usually
superfluous if the date is given in the references. Usually 'we wish to
thank' means 'we thank', 'has been shown to be' means 'is', and
'proved to be' means 'were'. 'It can be seen in Table 4 that' may be
omitted and '(Table 4)' placed after the statement. 'Because of the fact
that' may be cut to ' because'; 'concerning' may be replaced by 'on',
'therefore' and 'consequently' by 'so'. Indeed, so is a neglected word.
The following phrases may each be cut to one word:
actual truth as to whether weather conditions
all of completely full would appear
both of definitely proved absolute minimum
half of exactly true in an exhausted condition
quite unique first of all lose out on
very similar red in colour large in size
join together round in shape significant finding
As Strunk (8 ( 10)) writes, MAKE EVERY WORD COUNT.

Double hedging. Avoid repetition of the type 'may be probable',

'seems that .. . could be possible' and 'is supposed it might ... in
some cases'. Such hedging weakens discussion. Ifa writer is so unsure
that hedging is felt to be necessary, is he or she really ready to publish?
 12 COMMUNI C ATIN G IN SC IEN CE

One possibility seems to be that triple hedging may produce flabby

writing, as it has do ne in this sentence. If a hedging is later either
proved or dispro ved the early author can claim original mention, but
will not be admired for that.

(10) Pronouns. When you write it, this, which or they, are you sure the
meaning is plain? Readers may be unwilling to search back for the
meaning of 'it'. A pronoun deputizes (usually) for the nearest previous
noun of the same number (singular o r plural). If you have used a
pronoun for a more distant noun , perhaps the noun should be
repeated, as Summary is above [1 ( 4)). We may know what the author
means by 'seeds were placed in petri dishes which were then softened
in water' . However, if 'which' were to refer to something even more
distant than 'seeds', we might not know.
A sentence that starts with ' It' should be examined critically. It may
be that the sentence starts, as this o ne does, with an ' It' that does not
refer to a previo us noun, yet the reader momentarily expects it to do
that. So, turn ' It is believed that carbon dating gives' ... ' to 'Carbon
dating is believed to give ... '. Two other arguments against the
starting of a sentence with ' It' (needless words and distant references)
have been explained above (I (9) (10)).

Personal pronoun. An occasional ' I' need not be shunned. Indeed, ' I'
may be desirable to dispel doubt about who did something. If you
quote published results and then give yours, claim the latter. 'The
author' might seem to mean the other writer, not you . [See (11)
below.] Never, of course, write 'we ' for yourself, or use 'I' immodestly
or too often.
Personal pronouns should not appear in a summary because they will
make an abstractor's task difficult .

Tense, mood & voice

Authors usually write about their new work in the simple past tense: 'I
saw sparks' or 'sparks were seen' . D o not describe your results in the
compound past tense ; write 'were' not 'have been'. Other people's
work is variously reported. Many authors use the p resent tense; some
use the compound past tense, as in 'Nob has studied fossils'. A change
(11 ) in tense helps us to distinguish between your wo rk and his. Day [8 (5))
explains why the present tense is desirable in the reporting of published
work.
 WRITING A SC IENTIFI C PAPER 13

Working directions for a method are sometimes written in the impera-

tive mood. This is done partly because it makes the most direct style.
The passive voice. commonly used to describe results, sometimes
makes clumsy construction. Turn a passive phrase to direct style when
you can. For example, turn 'oxygen is needed for combustion' to
'combustion needs oxygen' . 'It is reported by Job' is better written as
'Job reports'. 'Excavation was involved in the project' should be ' the
project included excavation'. Sometimes a passive phrase may be
avoided by writing 'I'.
Never 'subject a patient to examination': examine him or her.

Choice or words
Beware of using words whose true meaning is not what you wish to
convey. Here are some words that deserve especial care.
case. I recommend that you read what Fowler, Partridge, Day or
other authority (Chapter Eight) has written about case. A sloppy misuse
is to make the word act as a pronoun - as in 'the above cases' - so that
the reader has to go back to find what the cases were. I have met cases
where I could not be sure to what the writer referred; did 'two cases'
mean two experiments, two animals or two observations on one
animal? Replace case, if you can , by a word that gives information, for
example 'this mineral' or 'Expt 8'. Or shorten the phrase, as in these
examples: in most c. (usually, mostly); in this c. (here); in all c.
(always); in no c. (never); in that c. (so); in the c. of (for, in); was the
c. (was true); the c. in question (this patient). ' In some c. this was the
c.' needs no comment.
different is used too often. If two methods were used, different is not
needed ; if the methods were not different there would not be two. In
' ... applied different torsions' replace different by various. From '4
different kinds' omit 'different'.
(12} due to and owing to. Due to often occurs where owing to would be
better. Due to has the sense of caused by; owing to has the sense of
because of. We may write 'the colour of the diamond was due to
impurities' but 'owing to impurities the diamond was coloured'. If 'Due
to' starts a sentence, that is probably wrong. Consider the sentence:
'Cardiac disease due to the use of drugs is not always fatal' . This
implies that the disease is caused by drugs. If due to be replaced by
o wing to, we have the opposite meaning: the disease is not always fatal,
because drugs are used. Commas make the meaning clear: 'Cardiac
disease, owing to the use of drugs, is not always fatal' .
1+ COMMUNICATING IN SCIENCE

Must we lose owing to owing to careless substitution by due to?

efficient describes processes whose efficiency can be measured. A
writer may mean effective. You may have devised a shaking machine
or a warning device. Can you determine that it is efficient? A
potentiometer-type power pack (for supplying desired voltages) was
described as 'efficient'. Engineers who read that such apparatus is
efficient, which it cannot be, may doubt the truth of other statements in
the paper.
fact. When you write fact, do you truly mean undisputed knowledge?
Effect, hypothesis, observation, value, result, phenomenon or finding
may be more modest. 'These facts' may even be changed to words that
give information, for example 'these similarities'. 'Due to the fact that'
is better written as 'Because'. 'In spite of the fact that my results were
negative' is bettered on several counts by 'Despite my finding no
effect'. Strunk (8 ( 10)) says, 'the fact that' should be revised out of
every sentence in which it occurs. Careful writers do not describe their
findings as facts. A 'fact' reported by an author may be contraindicated
by results from another. This often happens.
ftammable is preferred to inflammable. People sometimes take the
latter to mean it will not bum. The sequel could be disastrous.
minimal means lowest, smallest , and should not be written for small.
parameter is sometimes used unwisely. Variable might be safer.
varying means changing. The word is often used wrongly in place of
variable, varied or various. (1 (20). J Computations made with varying
formulas would have only little value , whereas those made with various
formulas might confirm one another.
washable. Does washable ink mean resistant to washing, or
removable?
which and that. If you are unsure about these words, recall the rule
'which describes, that defines'. Consider the phrases:
brown hens, which lay brown eggs, have yellow .. .
brown hens that lay brown eggs have yellow . . .
The first implies that brown hens lay brown eggs and also have the
yellow character. The second means that those particular brown hens
that lay brown eggs have it. Confirm your decision through the
commas; if they are needed, write 'which'. Another example follows .
Pronouns that cannot easily be identified with nouns are said to
dangle.
Dangling pronouns, which may be it, ones or these , are troublesome.
 WRITING A SCIENTIFIC PAPER IS

The first sentence defines the dangling pronoun and there are no
commas. The second sentence describes pronouns as troublesome.
Note the two commas, one before 'which', one before 'are' in the
second sentence; if the parenthetic remark between them be omitted
the sentence still makes sense: 'Dangling pronouns are troublesome'.
See also Homonyms, pp. 16 and 56.

Plain words. In general, use short rather than long words if they have
the same meaning. Write:
about, not approximately, of the order of or circa;
use, not utilize or employ (employ implies payment ; did you pay?)
have, not possess; enough , not sufficient;
Show may be better than demonstrate, disclose, exhibit or reveal.
However, do not eschew (avoid, reject, disdain, spurn, scorn) a grand
word if it conveys the meaning better than another: syrup is an aqueous
solution that cannot be called watery; expunge is more vigorous than
remove (1 (9)]; reveal is apt in Chapter Seven (1).
When you write the first words in the following list do you mean the
second , or vice versa? Brackets (parentheses); generally (usually); wire
(cable); if (when, whether); plug (socket).

Mathematical terms are often used for non-mathematical meanings,

which is undesirable if an ordinary word exists. For example, it is not
advisable to write centre (a mathematical point) if you mean middle, or
degree if you mean extent. For graphs, write filled symbol, not solid;
and unbroken or continuous, not solid, line. In mathematics, an area
has two dimensions, so use another word when you refer to more than
a surface. In scientific writing, negative is best reserved for minus and
for electric polarity; there are plenty of other words for none. 'No
response' is more scientific than 'response was negative'.
Avoid the unscientific use of ± for about or for with and without
(when + or 0 islneant), and do not use equals (=)for means. Certain
may sometimes be better than positive. Eight plus and 8 + are not so
scientific as > 8. (See also Chapter Four, pp. 47-8, on word processor
type faces.)

Elegant variation. English abounds in near synonyms: different words

that have almost the same meaning. Examples include begin, com-
mence, start, launch, originate, initiate, auspicate. Repetition of a
word within a sentence is considered to be bad style, which may be
 16 COMMUNICATING IN SCIENCE

avoidable with synonyms. However, in scientific writing, a synonym

should be used only if its meaning is clear. Repeat a word if the sense
so requires. There may be a case for a synonym where a technical word
might not be understood, but the writer must make it clear that the two
words mean the same thing. At the first use, some authors write both
words. [See The solidus, 1 (19) .) Repetition may sometimes be avoided
by the rewriting of a sentence. If a sentence has many and's, try
replacing one of them by then.

Homonyms. Many English words have more than one meaning. Where
possible, use a word that has only one meaning. Never, in one passage ,
use the same word for different meanings, as for example in 'Factor V
varied by a factor of 4' or 'For preservation one can can it'.
Normal has often been used as a trouble saver and has so many
meanings that it should be avoided where possible. Normal temperature
may mean 0 °C, 37 .4 °C or 273 K. Physiological saline is more descrip-
tive than normal saline. Normal should no longer be used to describe
solutions. The accepted form is mol 1- 1• Because mol 1- 1 is clumsy, is
not an adjective and is unsuitable for speech, many people prefer M,
mol/l or moVL.
The use of existing words for new meanings causes confusion,
especially to T. W. Fline . Chapter Six gives examples; here are
others.
Cell is overworked; cuvette is better in spectrophotometry.
Reduce has various meanings. Avoid the word or clarify it.
Figure is used for picture, pattern, diagram , shape, number, digit or
numeral, quantity or amount, price and value, as well as for calculate
and even for think. The reservation of figure for the first meaning
seems desirable. A number, such as 247, is composed of digits or
numerals; 24.7 mg is a quantity; 2.47 in a table is a value. The
abbreviation for ordinal 'Number' is 'no.'; cardinal 'number', used for
(13) quantities, should not be written 'no.'; write •... number of turns in
coil no. 6 .. .'.

(14) Foreign words. If you use foreign words when English words will
convey the meaning, you risk being accused·of affectation. You also
risk our failing to understand you because some of us do not know
many foreign expressions. Sometimes the grammar is faulty. Capita is
plural; hence, per capita means per heads. Did you realize that?
Strictly, media is plural, but in modern usage has often become
singular.
 WRITING A SCIENTIFIC PAPER 17

Conveyance of ideas without element or doubt

You may think this fuss over precision in the use of words ' is only
cosmetic'. But, if you prefer not to take rules of grammar and usage
seriously, may I plead as follows:?
First, convey your Message clearly. Dixon (8 (12)] writes in a breezy
style, yet his meaning is plain .
Secondly •. do not be conservative about names. We have discarded
vitriol, jar, fuming spirits of muriatic acid, candle power, probable
error. So let us drop a.m. and p.m., formalin, glycerine, h.p., pet.
ether, soda, carbolic. Is it fair to expect young people to learn archaic
names as well as systematic terms?
Thirdly, please do not accelerate the demise of hitherto useful
words. We have almost lost very, formerly a very useful word.
Consider that last phrase: did the second very affect the meaning? The
demise is hastened by such thoughtless uses as 'very unique' and 'very
level'; yet on occasion we need the word. (1 (6).) Surely, doubtless and
their synonyms no long.er mean without doubt; 'no doubt' they will
decline.
The meaning of quite is reversing: 'his method is quite good' now
means less good than good. Other reversals include release, which
formerly meant allow to go; now it is used for publish, that is push out,
as in 'Provisional data release'. Philip Norman (1980) uses the term
paranym for such opposites. In politics the use of paranyms is common,
but scientists should not use them.
'Fig. 5 clearly shows' is common. Had clearly been omitted, a reader
might believe the statement, but clearly alerts him, and the element of
persuasion makes him sceptical. Clearly is becoming a paranym.
'Plainly visible in the figure' has the same effect.
I appeal to you to use words with care.

Good workmanship endures

After you are elected a Nobel laureate, people will look at your early
papers. You may squirm with embarrassment if they discover that you
once thought accordance meant accord or that '4 times more than'
meant the same as 4 times as many as. [For argument, see Chapter Ex.]
Scientists who look back at their early papers feel (justifiable) pleasure
on finding good writing. I recommend you similarly to invest in the
future.
 18 COMMUNICATING IN SCIE N CE

You may think 'There are so many rules and pitfalls! H ow can I
remember them all?' At first you cannot . But , if you persist, you will
find that clear writing is a craft in which to take pleasure . The object of
a writer should be to convey information with minimal effort from the
reader. Although grammatical customs , like etiquette, are not all
logically defensible, if you ignore them you may obscure your meaning.

Language in flux
English is changing. That is desirable to meet changing needs, but it is
not desirable to lose meanings of useful words. When a new word is
needed, it seems better to make one than to add to the p lethora of
homonyms by taking an existing word. An example of such invention
is capacitor to replace condenser, which has other meanings. We need
a word for s.E.M. - without the 'error' connotation. Another could be
andor to avoid the algebraic and/or. But let us shun such horrors as
uni/ormization.
Certain changed usages are common and may become established.
Examples now occurring include: aliquot to mean any m easured
amount; a number were; these results suggest; ultraviolet light; under
the circumsta nces; de tergents and soap ; restructure d ; s ig nificant not
qualified by statistically; 100-volume H 20 2 ; heigh th; caustic, coronary,
medical and other adjectives used as nouns. The first few examples may
be acceptable. Othe rs are not. No doubt you could add to the list, but
do you find the trend agreeable?

Revision of the script must not be hurried

A writer's work m ay become so familiar that the author may be bored
when rereading it. The intense concentration needed for revision
(15} cannot be maintained for long. Therefore, I urge you to read only a
few pages of script at a time, and to
READ THEM SLO W LY.
What you have written should make sense , not only as you read it, but
when you read it aloud. Make it sound like intelligent conversation .
Where you pause, insert a stop. Ask friends to read the script to you.
Where they stumble, rewrite the passage .
Critical revision is m ore necessary than writers realize . Several of the
books on writing that are listed in Ch apter Eight contain errors and
lapses in style. There may be some in this book, too. So , BE VIGILANT.
 WRITING A SCIENTIFIC PAPER 19

Let your aim be to make such people as myself redundant. The species
is not yet endangered!
Extend your vigilance to repetition as well as to style and sense. In
some papers the Introduction and the Discussion contain similar
passages. In your script you may find parts of a method described in a
legend or in notes to a table as well as in Results. Avoid such repetition
if you can. If your Conclusion is repeated, as suggested elsewhere,
paraphrase it. Other parts, the comprehension of which is paramount,
may be said again differently, introduced perhaps by 'in other words'.
Use this device sparingly, however, or you will be in trouble, and so
shall I. For an example see 1 (3) with 1 (8), or 1 (1) and (2).
A great deal of needless repetition and verbiage is to be found, now
as in the past, in so very many published papers, yet it may be that the
authors were completely unaware of their unnecessary repetitions, a
possibility that can be adduced as yet one more good reason that
authors should have for asking a colleague both to read and to
comment on their scientific writings. You may like to write out the
previous sentence and, as you write, prune it to less than half. It can be
done. Please do not spoil it for others by marking the print .

Spelling
Some words have alternative spellings. For example, show, gray,
acknowledgement, disk, neutralize. If the choice is yours, use the
spelling that better represents the pronunciation. Some words are spelt
differently in the UK and the USA. A few publishers allow a writer to
use either, but the usage must be consistent. Check whether a drug
name that you are giving has different generic names in the USA and
UK.

Stops or punctuation , ; : . NP
The common stops may still be considered to comprise a hierarchy:
new paragraph, full stop (period or full point), colon, semicolon , dash,
comma. The dash and the colon do not always fit into the hierarchy.

A new paragraph (NP) , in general, denotes a change of subject. In

good prose, one notion leads to another, which makes this rule difficult
(16) to apply. Because short paragraphs can look irritating, long ones
boring, skilful compromise may be needed. If you want each new
20 COMMUNICATING IN SCIENCE

paragraph in your published paper to show, you MUST either indicate

each one by indention or mark it on your script. [Seep. xii.]

Colon. The phrase before a colon is general; the phrase (or phrases)
after it is (are) particular. Therefore, no sentence should contain two
colons. Traditionally a colon does not end a sentence, but a full stop
does, so a colon should not occur after a phrase 'that refers to more
than one sentence each having its own full stop. For example, if 'as
follows' is followed by complete sentences, use a full stop, not a colon,
after 'follows'. The following type of confusion is submitted to editors.
The solution contained: glucose: 2 g, NaCl: 3 g and urea: 4 g.
The piece is better written as follows .
The solution contained: glucose, 2 g; NaCl, 3 g; and urea, 4 g.
Similarly ' ... conditions (time: 25 min, current: 2 A)' should have its
comma replaced by a semicolon and its colons by commas.
Some people would banish the colo.n; yet it can have a real use.
Please help it to survive and to keep its value - greater than that of the
comma, less than that of a full stop.

The semicolon differs from the colon in various ways: it signals a

shorter pause; it separates items in a list; it joins closely related clauses.

The comma signals a short pause . Use commas cannily, and more often
than some writers use them , to prevent over-reading - as in the next
sentence. Where two adjacent nouns belong in different clauses
separation should be achieved with a comma - for example after
'clauses'.
Children used to be taught not to put a comma before and. The logic
is as follows. We may mean a and band c, but we write a, band c. One
and is replaced by a comma, but no comma is needed before the
retained and. In the phrases 'chalk and clay' and 'tides were measured
and recorded' no comma is needed. The and is the joining variety and
could be represented by& . There is another kind of and: 'the cats were
fed on meat and worms were given to the fish' . Your probable
hesitation could have been avoided if a comma had appeared after
'meat'. Here is another example:' ... is dissolved in 5 mo! NaOH 1- 1
and 2 mol KOH 1- 1 is added'. And another: 'The chairman was head
of the physics Jab and the principal of the maths lab was elected
vice-chairman'. So, when two phrases are linked by and a comma is
needed to show that they are two phrases. Curiously, the superstition
does not ban the comma before or, or before but.
 WRITING A SCIENTIFIC PAPER 21

A comma is needed before an and that separates negative and

positive notions, as in 'Do not write too much between full stops and
present the information in small packets for easy understanding' and in
'The alloy is made by adding Sn to Pb and Zn is rigorously excluded'.
A comma can change or even reverse a meaning. 'The sky is not
black, as Zob claimed' means Zob claimed the sky is not black.
Without the comma the phrase means Zob claimed the sky is black.
Consider these two sentences. (a) Scientists, who are honest, report
what they observe. (b) Scientists who are honest report what they
observe. Sentence (a) defines scientists as being honest. History shows
that a few have been dishonest, so sentence (b) is preferred.
Commas are often used in pairs, as here, to enclose a parenthetical
remark. The latter is treated at 1 (18).
A comma can change the meaning of a word. In 'However he added
... ', However means 'in what way'. In 'However, he added ... ',
However means 'But'.

(17) Dashes. The four symbols, hyphen(-), en rule(-), minus ( - ) and em

rule(-), are represented on a typewriter and word processor by one
sign popularly called a dash. Nowadays, a short dash (en rule) with a
space each side - as here - usually replaces the long dash- the em
rule-but the meaning remains the same. The meanings of the hyphen
and the dash are opposite: the hyphen joins words or pieces of a word;
the em rule pushes words apart with a pause longer than that signalled
by a comma.
The hyphen and the dash are distressingly often confused even by
educated people and trained typists. I appeal to you to persuade your
colleagues to preserve the identities of these useful signs.
The en rule (short dash) has various uses, including that in 1978-89.
Spaces are not needed. Write 'from 2 to 28' not 'from 2-28'. Solutions
are sometimes described thus:
hydrochloric acid-sodium chloride

ethanol-ethanoic acid-water

A void confusion by writing 'mixture of hydrochloric acid and sodium

chloride' or 'mixture of ethanol, ethanoic acid and water (3: 1 : 10)'.

The hyphen has many uses. Because the rules cannot be condensed to a
few lines I shall do no more than offer suggestions and examples.
 22 COMMUNICATING IN SCIENCE

A hyphen joins words to make adjectives: if a hypothesis is well

known it is a well-known hypothesis ; an example is the all-or-none
hypothesis. Do not join adverbs that end in -ly. Write 'vitamin-
deficient and rarely eaten food'; also write 'X-ray-induced and chemi-
cally induced mutations'. A hyphen can change a meaning: a large
impulse counter is not the same as a large-impulse counter; 5 day
sessions differs from 5-day sessions.
Hyphens may be used to join nouns to make an adjective (vanadium-
steel pin, noun-adjective phrase), but if this be done with multiple
nouns the result is clumsy, as in negative-particle-analysis procedure
and Lipon-8B-column chromatography. Such phrases should be
turned; e .g. 'ablacil-C-induced effect' should be 'effect induced by
ablacil C' .
Consult a recently issued and well-produced catalogue for help with
hyphens in chemical names. Catalogues from BDH (8 (20)] have been
useful.
Link a prefix (non-, post- , etc.) to a noun by a hyphen. T. W. Fline
(p. xiv] are then warned to look in the dictionary for the noun instead
of searching under n or p. Postparturition might confuse him. Insertion
of a hyphen may aid the pronunciation of words such as sub-unit,
co-operate and co-worker.
If you invent a new name, for example Q effect or HK thesons, do
not use hyphens. 'Copper compound' should not be joined by a
hyphen; nor should 't test'.

( 18) Parentheses is the name for round brackets ( ) . Parenthesis also means
an aside or explanation between two commas, dashes or parentheses
(as here) within a sentence. If such a parenthetic remark is left out, the
sentence should still make sense and be grammatically complete. If the
remark is a complete sentence give it a capital letter and its own full
stop, surround the whole with parentheses or square brackets, and
place it after, not within, its parent sentence. [4 (1).] Although ' (Table
2)' may appear within a sentence, it is better for a longer° phrase such as
' (See for example Table 2.)' to be treated as a sentence. If the
parenthesis is only part of a sentence, the full stop goes outside.

(19) The solidus (diagonal , slash) is used to mean 'per' as in wires/cable.

Confusion may arise if the diagonal is used for other meanings, such as
dates or contractions. A solidus may make 'response/dose' better than
dose- response.
 WRITING A SCIENTIFIC PAPER 23

A solidus brings a mathematical formula into one line, thus:

(a+ b2 )13p. Fractions going through two lines are ugly.
The solidus should not be used for plus, and, with or meaning other
than per except after careful thought .

Initial Letters. Avoid starting a sentence with a lower-case abbreviation ,

such asp- or a-, or with van der, if you can. A sentence should not start
with a numeral. If you find such a sentence , either rewrite it or spell out
the number. This practice is especially important when the previous
sentence ends with a numeral , symbol o r abbreviation: change ' ...
found in 1989. 73 µg of iodine was . . . ' to •. . . found in 1989. Iodine
(73 µg) was . . .'.

Abbrevns & cntrctns

The smallest symbol in the printer's fount, the dot(.), has many
meanings. Its second best-known function is to indicate abbreviations.
The abbreviative presence - or the absence - of the dot evokes more
editorial adrenaline than does that of any of the larger symbols.
So here I indicate practices but do not dare to instruct.
Some publishers omit the dot from Expt, Mr and Dr as well as from
other contractions that include the first and last letters of the word.
Some words (Miss, log, bus) have been with us so long that few people
give them a stop. Units (g, mm , F, min, h) take no full stop (period)
and plurals no 's'. No dot is needed in acronyms: laser, radar.
A few abbreviations are ugly. Viz. saves only two cllaracters from
namely, and c. saves three from about. Approx., too, is bettered by
about. Cf., which means compare not see, may usually be deleted.
The Biochemical Journal, in its instructions to authors [8 (2)], gives a
list of abbreviations useful to all scientists.
Abbreviations of long terms, or of names of materials to which you
often refer, are best collected on a page by themselves. The names
should be given in full in the Summary.

Headings or captions
As a novelist uses dialogue to make a page look interesting, so a
scientific writer uses headings and subheads. They help to make a
paper readable, and guide inquirers to parts they want to read again.
Use many. Write subheads for the Discussion; they can be truly
helpful.
2+ CO MM UNICAT ING IN SCIENCE

A heading should contain a noun. A lo ne adjective sho uld not be

used; add "part' to Experimental.

Make your headings work. Be cunning. Perhaps you have used the
'in-other-words' device in the text but still desire emphasis. T ry to
include, in a heading, the notion that needs emphasis, but use different
expressions in heading and text. (See I (15) on revision , o r 4 (2).)
Repetition of a heading in the text is undesirable. A heading may ask a
question [Ex (1)). The Introduction to a normal paper should not need
that word as heading, which is as superfluous as is 'Notice' on an
obvious notice. However , if you can put information into a heading for
the Introduction, that could be useful. For example , can you describe
your problem in words different from those in the Title?

Tables
Study the tables in the journal of your choice; then conform with the
style so that the ed itor does not have to make rearrangements.
A table needs a title, probably supplemented by an explanation. The
a
heading to each column usually includes units , so that each entry is
number. If your quantities are large or small, use the prefixes M, m, p,
etc. Avoid using x io- 3 in a heading because a reader may wonder -
have you divided by 103 or is it still to be done?
Indigestible tables with many or cumbersome values deter readers.
Trim the values even if you risk losing an occasional significant digit.
Methods exist for calculating significant digits. One digit more than is
meaningful does no good whatever. Indeed, a writer who presents
non-significant digits is displaying a lack of understanding. The S.E.M.
should have no more digits after the decimal than does the mean.
Should the argument require results from several experiments, and
they cannot be condensed.or pooled [see 1 (5)], consider dividing them
into two or more tables.
When measures of variation are given there must be no doubt about
their meaning. Common useful measures are the following .
The standard deviation (s.o.) gives an estimate of the spread of a
measurable variable. The dimensions o r units (pascals, mg/I , ... ) are
those of the variable being measured.
The standard error of the mean (s.E.M.) gives an estimate of the
precision of the mean.
The coefficient of variation (c. v.) is dimensionless and is useful in
comparisons between populations. The c.v. is usually the s .o .
expressed as a percentage of the mean.
 WRITING A SC IENTIF I C P AP ER 25

Unfortunately s.o. and s.E. have sometimes been used interchangea-

bly. In one paper submitted, the sta ndard erro r of the mean was so
described several times, but the formula, given three times (in text,
table and note), was that for s .o. You may often encounter s.E.
without the M . But common p ractice does not make it good practice.
For clarity's sake include the M .

Illustrations
When an experiment provides many observations they may be better
given in a graph than as a table. For many people a diagram is easier to
grasp and to re member than is a table . The same information is not
usually allowed to appear in both forms.
The horizontal co-ordinate of a graph represents what we select
(time, weight, frequency ... ) and the vertical co-ordinate what we
measure. If the origin of an axis is not zero, indicate this by a break in
the line. A graph needs an indication of precision , such as an estimate
of confidence limits, or symbols of a size to indicate the s.E.M.

Units & quantities

Use SI units. Some old units (e.g. bar , calorie, mm Hg) may survive for
a time, but, when old units are used , SI units should be given too.
Modem units go up and down in steps of 1000. Avoid other steps if
you can. The angstrom is redundant - and mispronounced! Concentra-
tions other than mo11- 1 are expressed as parts per thousand o r per
million (p. p .m. ) . Concentrations should no longer be expressed as o/o.
What is still sometimes called molecular weight is a ratio rather than
a weight. This ratio is printed M ,. Dixon (1983) explains why relative
molecular mass is a better term than molecular weight. The inferior, is
typographically unsatisfactory, which may be one reason for the
unpopularity of M ,. Another m~y be the resemblance to Mr (mister).
See, for example, what looks like 'Mr. Dixon' above.
Make you r units unambiguous. If you 'add 2 mo11- 1 HCI', either say
how much HCI you added or that you 'made.the solution 2 mol 1- 1
with respect to HCI'. Better still , write 2 mot HCI 1- 1 •
T imes should be given on the 24-h system, e.g. 08 h 30. Dates are
printed without stops (06 April 1990) or with hyphens (06-iv-1990).
Logically we should proceed from large to small units (year, month ,
day, hour, min . . . , thus 1990-04-06 ... ) as astronomers do and as we
do with weights and measures. T his logical system is described in
British Standard 4795 (8 (19)) and in IS0-2014; 1975.
 26 COMMUNICATING IN SCIENCE

When you offer a series of values, do not repeat the units. Write '3
and 4 g' not 3 g and 4 g. Another kind of repetition should also be
avoided. From 'the percentage was 8o/o' omit % or write 'the propor-
tion was 8°/o •. Write 'the pH was 8' , and 'the voltage was 8' or state that
the e .m.f. was 8 V . People do not write that the ohmage was 8 ohms or
that the hourage was 14:00 h.
Units named after people are spelt without capitals but symbols for
such units do have capitals: thus, watt and W, joule and J, pascal and
Pa . Write lower case k for kilo.
Because I, 1 and I may be confused, especially in scripts and in
sanserif (I, 1, I), Lis better than I for litre. For an entertaining history of
the origin of L , see Wolner (1975).

Good sense
R ead other people's writings critically. Even edit them - but not in
library books! Improvements can usually be visualized in, and words
may be deleted from , ·MOST papers (as well as from Directives to
Authors in some journals!) without effect on the meaning. You may
obtain amusement by collecting oddities, and this will help you to be
vigilant in avoiding them in your own writing. Here are some: absent
in the solution (write from); molar CuS04 ,SH 20 ; the kind gift; at
(20) varying temperatures in a thermostat (write various); o ur records show
that you do not exist; pressure of space; specimens were stored in a
refrigerator wrapped in Al foil; beam from a Philips machine filtered
through 1 mm Cu; they were in fact not artifacts. The title of a paper
seen in a medical journal: 'Prevention of recurrent sudden death.'
More desirable is the collecting of 'good stuff' (Herbert, 8 {11 )).
When you meet good writing, study it . Try to analyse your liking for it.
Then emulate it. Through attentive reading you can enlarge your
vocabulary . This will help you to overcome a principal difficulty in
writing: finding the right word. When you have this difficulty (quan-
dary, problem , doubt, perplexity. dilemma ... ) , write various words
(as here). Later, you may choose (select, reject, pick, exclude ... )
suitably. If no word fits, consult Roget's Thesaurus (8 {17)]. If you meet
a 'new' word you like, look it up in a d ictionary before you use it. Then
you will be unlikely to write a howler such as ' two of the six quintiles'.

Emotion & modesty In scientific writing

May I suggest that you avoid emotion in scientific papers? 'Great
importance', 'significant conclusions' and similar expressions should be
 WRITING A SCIENTIFIC PAPER 27

restrained. If you are tempted to write 'this most interesting result' ask
yourself 'To whom is it most interesting?' Strunk (8 (10)) writes
'Instead of announcing that what you are about to tell is interesting,
make it so' .
A system, e.g. blood clotting, may have several essential com-
ponents. An experimenter says that the component he studied is
important. What is he telling us? 'Important' is a superfluous term for
an 'essential' component.
You must mention your publications as necessary, but do not let the
Bibliography look like a personal history.
If Lob published first , write 'my result agrees with that of Lob' not
'Lob confirmed my result'.
An author who writes 'we were surprised' or 'unexpected result'
admits lack of knowledge, for those who know all can predict all.
Humility is good but may not be the author's intent.
(21 ) Promises should be offered only sparingly. Authors who write that
an idea will be investigated may be warning you off 'their' territory.
Why omit 'other' from 'human and other animals'?
If you list arguments as 'first, . . . second(ly), ... 'avoid calling the
last 'finally' . You can rarely be certain the subject is closed.
Of course we hope - but privately, not in a scientific paper.

The three saddest words in the language

'The Editor regrets .. .'. So what should you do?
First, do nothing, or at least nothing impulsive.
Second, take note of the reasons the Editor gives for non-acceptance
and consider taking his or her advice. Most editors are kindly people.
Third, pay attention to what the refere.es say. [Reviewers in the
USA.] They may have spent much time on their reports.
Fourth, read this chapter again. Read it slowly; you skimmed it too
quickly last time.
Fifth, rewrite the paper. Then show it to two critical colleagues.
Chapter Two
Before you lecture or talk to us,
please read this

Craftsmanship in speaking at meetings

Most scientists attend conferences and listen to talks by other scientists.
Occasionally, a speaker so enthrals listeners that they enjoy hearing
about the work and they listen intently. On the other hand , a speaker
may be so dull , ill-prepared or inaudible that listeners fail to follow
what is said.
Between such speakers are others, most of whom could deliver a
better talk if they were helped. Some speakers may not realize that
they could improve their performance. Many conference attenders
have told me that they have suffered while striving to listen to such
speakers. Because I have suffered too, I have been inspired to write
this chapter. Books on lecturing exist, but scientists are too busy to
read them. Perhaps they can find time to look at this chapter or at least
to read about empty words and visual aids.

Direct style
Part of what I have written is in the imperative. This is because the
imperative allows the most vigorous and most readable style. I write as
though I am speaking to you because one day I may be listening to you .

Bad manners
If a scientist gives a talk without taking care to make it easily
understandable, that is bad manners. Sometimes I think that a person
other than the head of the Unit should be giving the talk.
The idea expressed in the first sentences in this chapter is so
important that I shall repeat it but in a different form . Moreover, you
may meet the idea yet again elsewhere. High you may be on the status
28
 BEFORE YOU LECTURE OR TALK TO US, READ THIS 29

ladder, but, if we impatient listeners on lower rungs cannot understand

you, we shall not esteem you and your communication will be wasted.
On the other hand, if you enlighten us, your stature will increase.

Title of your talk

.
Look at lists of talks; some titles tell us what a talk will be about, some
do not. Ensure that yours does. Try it on colleagues. Avoid standing up
at the start of your talk to tell us that the title given in the programme
is inappropriate.

How to begin
Listeners may not hear the first sentence of a talk; they are 'tuning in' .
So do not start with crucial information. If you have to say 'it's an
honour to be invited ... ', let that be your opening, but limit it to one
sentence. If you do not need that opening, look to the back of the room
and ask 'Can you hear me?' I have done that; it works.
Now begin.

What to talk about

Please tell us
1, why you did this work;
2, how you did it;
3, what you found;
4, what you think it means.
Thank your audience ( < 6 words); then stop.
A scientific talk should not take the form of a written paper. You are
telling us of your recent discovery, so the first three parts make up the
necessary prelude to your main Message in part 4. In a written paper, a
scientist must so describe the method that another can repeat the
experiments. In a short talk, we must assume that your method was
suitable; a description of the principle of the method should suffice.
Anybody who wishes to pursue your experiments can speak to you
afterwards.
Part 3 should take most of your allotted time. You have to tell us
enough results to support your conclusion, but you must simplify them.
The results must not be cluttered with statistical details. Just tell us
the averages (for example from your experiments with and without
30 COMMUNICATING IN SCIENCE

treatments) and their statistical significance; we have to trust that your

calculations are correct. A graph is more easily understood than is any
but the simplest table; the more results you have the more complicated
is the table, but the better the graph.
Now tell us your Conclusion (part 4). This is what we have come to
hear. So you must do all you can to convey this part clearly. Let there
be no distraction : tum off the projector; do not rustle your notes.

Verbal delivery
Let us consider how you can make yourself heard and understood.
A good speaker looks at the audience, not at the furniture, but does
not stare at any one person. Hold up your head and speak to people at
the back of the room. Then those at the front should hear too.
Do not read your 'talk' . If you read, you may speak too fast or in a
monotone; you will look down, compress your lungs and perhaps
become inaudible. Speaking to your stomach will do it no good. If you
have to read a passage, hold the book or paper high.

Notes as prompter
Seep. 2, note 3 about 'reservoirs'. From such reservoirs, pre pare
notes; if they help, use them. There is nothing shameful about that.
Notes are better written as headings than as sentences: then you can
refer to them at a glance. Lettering should be large, clear and well
spaced , in lower case , not capitals, so that you can read them without a
lectern lamp. The lamp distracts listeners and casts unflattering light on
your face . Make the note on cards the area of this page so that you can
carry them if you walk to the wallboard .• When you think the notes are
ready, go through them with a red crayon and ruthlessly slash out
non-essentials. Number the cards. My (paper) notes were once blown
to the floor by the projector's fan. How glad I was that they were
numbered. Now I use cards.
On the top card, write a list or items to take into the lecture hall:
clock, mask for the projector, coloured markers, pen, books,
exhibits . . . .
If you have slides, apparatus, specimens, transparencies to exhibit ,
signal them with numbers on your notes - in colour. The audience will

• The blackboard is now often white or green.

 BEFORE YOU LECTURE OR TALK TO US, READ THIS 31

not appreciate your: 'Oh yes; I should have shown this earlier'.
Numbers of the reminders should tally with numbers written on the
items.
Some speakers, not knowing what to do with their hands, put them
into their pockets. This is inelegant. Notes help, by occupying your
hands.

Stage fright
If you have delivered many talks or lectures, please skip this section.
If you are a novice, you may wonder how to overcome nervousness.
Most people are nervous before or during their first public talk. With
practice, the nervousness lessens. One man's knees shook violently as
he stood on the platform but he concealed the shake by holding the
table. Several talks later he felt only a trace of nervousness before, and
none during, his talks. People need practice, and the smaller their first
audience the better. This is just one of the reasons why every
laboratory or institute should have a tea club at which a researcher tells
colleagues what he or she is doing. If your institute does not have a
club, perhaps you can found one . [I (l).]
Starting with a large audience can be devastating, as the following
incident shows. The television broadcast had begun. The next speaker
was awaiting his turn. Suddenly he said 'Oh, my car!' and ran out. The
producer assumed the man had remembered parking his car badly. He
did not return . . ..
One precaution to take against stage fright is to observe continual
foresight against your notes going astray. You could [not many would]
make a copy to keep in a pocket. The knowledge that all is in order
helps you to relax.
If you let it be seen that you have not prepared your talk, and you
fail to speak coherently, the audience will not like you. However, if the
audience senses that you have taken trouble but have stage fright, they
will be sympathetic. A sensing of such sympathy encourages a speaker,
and the nervousness diminishes.

Verbal style
We, your listeners, prefer that you 'talk' to us rather than deliver a
formal speech . This calls for a simple style. Here are suggestions.
In general, use short sentences. But if they are all short, delivery will
be jerky. A pleasing sentence usually has two main verbs. but an
32 COMMUNICATING IN SCIENCE

occasional very short sentence may be used for emphasis. (See the next
paragraph.) If an interpreter is present, remember that simple sen-
tences can be translated more easily than complex ones.
11 Please finish every sentence.

lmportanceof deliberate speech

Speak so slowly that it seems almost absurd. There are good reasons.
When we stand, heart throbbing, before an audience, adrenaline
speeds us up, and we do not realize how fast we speak. During a
lecture I was giving about speaking, I asked the audience to observe its
pulse rate. The average was 65 beats per min; mine was 130. This was
taken as evidence of my speed-up. When the experiment was repeated
a dozen lectures later, my rate was only 75 beats per min. I had become
adapted to lecturing and had to give a false value to make my point.
A maximal rate for speech is 100 words per min. So a 10-min talk
should be kept to below 1000 words.
Rehearse your talk to a colleague, to a tape recorder, or to a video
camera. Use a clock. Remember, the person who exceeds the allotted
time is a thief and is also unwise, because a strict chairperson may call a
halt before the Conclusion has been uttered. Ask your critic to time
you and to tell you of your distracting mannerisms.
If you've not heard yourself before, you may be shocked when you
hear the playback , as most of us were. Do you swallow the ends of
words, as English-speaking people often do? Note whether you speak
in an exPLOsive manner; this is disTURbing.

Utter no redundant words

If you are to speak slowly, yet have much to say, how can you resolve
the dilemma? Contribute to a solution by uttering necessary words
only; that is, aim at a high signal-to-noise ratio. Waste no words in
saying you have not time to tell everything. Such non-information
irritates the audience .
Omit: it has to be pointed out that; as a matter of fact; it only
remains to be said that ; the whole point is .... We have not travelled
all this way to listen to such 'empty words'.
Avoid tautology [repeating, but with another word): pooled to-
gether; foot pedal; circular disk; time clock; supplemented with
additional salt; still remains; knots per hour. Obvious? Yes, but these
 BEFORE YOU LECTURE OR TALK TO US, READ THIS 33

double sayings do creep in. Whenever you see the offer of a ' free gift',
renew your vow never to tautologize.

Andermanship makes us squirm in our seats

This striving for economy also means do not start sentences with
'Anderm', the most irritating non-word ever misfangled. When you
listen to an interview on television, compare the interviewer's speech
with that of the untrained interviewee. If you find yourself counting the
times the latter says 'you know', 'that sort of thing' or 'basically'
perhaps that will stimulate you to listen to yourself on a tape recorder,
identify your personal in-fill word (or non-word), then train yourself to
forgo it. That is difficult, but it can be done; it must be done if you are
to become an Eloquent Scientist.
~ I entreat you to read that last paragraph again. And then at least
once more.
Speak deliberately; carefully choose good words. You can then offer
more information per unit time than can those who talk fast and join
phrases intoalmostinterminablesentencespaddedwith empty words and
.
noise.

Hackneyed phrases
Cliches, such as 'tip of the iceberg', should be 'avoided like the plague',
banished 'for ever and a day' 'and then some'. Can you recall a sillier
phrase than 'corner of the globe'? The foreigner who thought redox
meant red ox would be puzzled by 'quantum leap' used outside atomic
physics.

Pronunciation
Think ofT. W. Fline while you speak. [Page xiv.) Articulate each
word distinct I y for them, and pronounce even the unaccented
syllables at ends of words. Then your own nationals will hear you too.
(See also pp. 59-60 on speaking abroad.)
Technical words should be pronounced unmistakably. For example,
pronounce they in methyl and in benzyl as in 'by', even if it is not your
local custom to do so.
People from Britain or from North America, when speaking in the
other place, should remember that not only does pronunciation differ,
34- COMMUNICATING IN SCIENCE

but different words may be used. [Fume cupboard, hood; earth,

ground ... .J Therefore, misunderstanding must be forestalled.
If English is not your native language, read Chapter Five.

Keywords
Your talk may depend upon our hearing a particular word. That word
may be drowned by a local cough or scrape. Then we lose the thread.
Therefore, write such KEY wo Ros on the wallboard large enough for
me to read them at the back. Point to each as needed. The shorter the
word the more important this gesture. Do you see why? On your notes,
write each key word 1 N co LO u R.

Partial deafness
Perhaps as many as one in four of your audience has a hearing defect
and may not even realize it. The reduced hearing is not evenly spread
across the auditory 'spectrum': usually the loss is more severe at high
than at low frequencies , although there may be selective loss at mid
frequencies. So an amplifier may not help. What does help is clear
enunciation. That cannot be given in rapid speech; so here is good
reason for you to
speak slowly .
Partially deaf persons understand best if they see a speaker's lips. So
stand in the light and look at the audience. Be tidy, so that you are easy
to look at.
Auditory impulses go from the cochlea to the brain. They are then
converted into meaning. You can do experiments to show that the
conversion takes time. The rate of this conversion or cognition
becomes slower with age.

Words
Use short words where you can: start, not commence; try, not
endeavour; often, not frequently .... An uncommon word may
express a writer's meaning exactly. Examples occur in this book. A
reader can consult a dictionary about an unfamiliar word. A listener
cannot do that , so a speaker should avoid such words .
Vogue words, too, should be avoided, unless they are really suitable.
Examples are: focus, restructured , breakthrough, situation. Some
vogue words do not even have the meaning speakers ascribe to the1n.
 BEFORE YOU LECTURE OR TALK TO US, READ THIS 35

Perspective , for example, does not - o r at least did not - mean

viewpoint; it means two-dimensional representation of three-
dimensional objects. Before you say perspective, please consult a
dictionary; then say viewpoint if that is what you mean.
Infinitely variable should not be uttered for continuously variable if
one or both limits are finite .
Constantly is 'constantly' misused (as it is here) to mean no more
than often. Continually, repeatedly, regularly or even sometimes, may
better represent the desired meaning. Reserve constant for unchanging.
Say constantly changing only if you mean just that. Only say invariably
if you mean always; even better, say always. A speaker who says that
bad weather constantly [or invariably] interfered with observations
means annoyingly often , and he implies that the Law of Exasperation
was at work.
Relatively and comparatively should be used only if things are to be
compared. Alone, these words have limited meaning. So has •a fraction
of'.
While should be restricted to its temporal meaning; try whereas,
although or even and. Similarly, because sometimes betters since. Did
an author really mean 'A began an experiment while B finished it'?
When you speak. use no foreign words for words that exist in
English, even though you might use them in writing. (1 (14).]
Say (and write!) less material but fewer things.

Contemporary terms
Be up to date over units and technical terms. Young listeners who have
to ' translate' archaic names may 'lose' your next sentence. Such terms
as condenser (for capacitor) or normal solutions may not be familiar to
those who have recently left school.

Wolf words
People needing a word having the sense of very, yet knowing that word
to be dying, resort to totally or extremely. These words once had the
sense of ultimate, utmost, the highest possible, but such words are
losing that meaning: for example, 'utterly vital' has come to mean no
more than 'necessary'. If a speaker exaggerates, listeners will find him
out and be suspicious of his other statements. Aesop's tragic story of
the boy who cried 'Wolf!' is still valid .
36 COMMUNICATING IN SCIENCE

Do you think there is a difference between

completely empty and empty.
absolutely none and none, or
much more and more?
Can you truly say that anything other than perhaps the speed of light
(electromagnetic radiation) is extremely fast?

Pronouns it & I
Speakers may say ' it' as a stand-in for a word so distant that we cannot
recollect it. Be safe: repeat it. [Do you see the point? The it should
have been the word.] Listeners cannot ' refer back' as readers can.
The word ' I' need not be avoided altogether. Try to convey your
enthusiasm about your work to us. An occasional 'I' helps, although
too many are bad. Of course, you will not call yourself 'we' - unless
you are royalty, an editor, pregnant, or the spokesperson of a team.
Sometimes a speaker says 'we' to mean 'you and I' .

Teleology
In a scientific talk on television the speaker may say 'the bird has
developed a long tongue so that it can reach nectar', or 'bright colours
have evolved in flowers to make them attract insects'. The speaker
would be wiser to say 'because the insect .. . it can reach' or 'because
the flower is coloured it attracts ... '.
There seems to be no evidence that DNA can think teleologically; if
it could, it would have provided scientists with the gift of eloquence.

Diversions
For various reasons, listeners find it hard to concentrate, and they are
easily diverted. If there is an interruption beyond your control and
people look away from you, tell us 'I'll say that again'. Your courage
will be admired. Some diversions may even be induced by the speaker
him- or herself. Therefore, eschew irritating mannerisms, magnilo-
quent words and complex sentences.

Keep it up
Some speakers start well but gradually lower their voice. On your notes
write an occasional reminder to speak up.
 BEFORE YOU LECTURE OR TALK TO US, READ THIS 37

Courtesy requires attention to detail

You may think I am a critical fusspot and that these details do not
interest ordinary people. But we are not ordinary people; we are
special people. You would not be reading about speaking were you not
a Special Person. Therefore I hope you will campaign with me for
better scientific talks.

Questions from the audience

You may be asked about experimental details or about your Conclu-
sion during the discussion. Anticipate critical questions, and be pre-
pared to answer them succinctly; yes, succinctly because most of the
audience may not be interested in one questioner's problem.

Visual aids
If you use a projector, let your tables be truly simple. Never display
more results than the argument needs. Tables from published papers
may not be suitable for the screen and may provoke resentment. So
make tables especially for the occasion. Then project the most crowded
in an empty lecture room. If you cannot read a display from the back of
the room, remake the table. If an overloaded table is shown on the
screen, or on the wallboard, listeners wonder whether to study the
table or listen to the speaker. They may attempt both and in their
confusion do neither, so the speaker might have achieved more without
the display! Do not think that saying 'You need only look at the bit in
the bottom right-hand corner' helps us to read an illegible slide.
If you make slides, read what Norris (1978) writes about their
preparation. His essay is excellent.
If an indicator of magnification is needed, show a measured bar. On
the screen, 'x 1000' becomes meaningless.
Before you use an episcope, or the epi part of an epidiascope, in
public, check whether what it projects is satisfactory as seen from the
back of the hall.
A hand-held illuminated arrow should be switched off when not
needed. Some speakers wave it about distractingly, or even, unknow-
ingly, set it down pointing at the audience. Switch off the projector
immediately you have used it , even though you may need it again, and
turn out the lights. The audience should pay attention to what you are
saying now, not to what you displayed minutes ago; and they should
38 COMMUNICATING IN SCIENCE

not be dazzled by a bright empty screen. Write reminders on your

notes.
Some speakers project notes on to the screen, then talk around
them. This is undesirable on three counts: the room has to be
darkened; the audience reads ahead; and the speaker tends to talk to
the screen. These objections vanish if an overhead projector is used,
together with a mask, although speakers must remember to address the
audience not the projector.

Overhead projector
An overhead projector has advantages over a slide projector (dia-
scope). (Appendix, p. 40.)
Before you prepare transparencies, take a white card and draw lines,
about 20 mm apart, across it. If you lay a transparent film on the card,
the lines will help you to arrange your words neatly. As a compositor
says: it's the arrangement of the white spaces that's important. Prepare
your transparencies with permanent ink , then use suitable dense
water-soluble ink for numbering the transparencies, in accordance with
your notes, as well as for marking during your talk. After the talk or
lecture, your marks and the number may be washed off, but the
original writing remains for re-use. Write with bold clear lettering, NOT
IN CA PIT A LS. [4 (3).] Typewritten characters are too small. We are
not entertained by the.remark 'I suppose you can't see these numbers'.
Separate the slippery transparencies between sheets of white paper:
(a) to lessen sliding; (b) so that you can read the top one.
Before you project, place a mask of card over the display; then move
the mask to uncover what you want the audience to see as you speak.
Do not first display the whole, then cover it; that is discourteous.
In a very large room an overhead projector may not be suitable.
Another system must then be used.

Microphone & loudspeakers

Should you use a speech amplifier? If you have a strong voice and will
speak up- no. Even the best systems give some distortion; and your
freedom may be restricted . If you have to face the screen or wallboard,
however, an amplifier may be advisable.
 BEFORE YOU LECTURE OR TALK TO US, READ THIS 39

Light relief or humorous remarks

Should you tell a funny story? Yes, if it be relevant, brief, witty and
dignified. Not otherwise. Favourable response from an audience dispels
nervousness and helps you to speak better. And we shall listen alertly
in the hope of more fun. Perhaps you can recall an incident from a
previous lecture, or you may have seen an oddity such as 'Drugs were
given to patients dissolved in alcohol.' Bear in mind that feeble or
crude jokes will do no good and may damage a lecturer's image.
Having no response from what the lecturer thinks is funny can be
devastating.

Television & videotape

Speaking for television needs especial care because television loud-
speakers are not usually hi fi. Before the talk, a speaker might ask the
producer not to show a close-up in which the face fills the screen. If it
does, viewers may study details of the face instead of listening.

This chapter does not tell everything

I have made suggestions. You can probably make others. There is
space for them at the back of this book.
You may like to read Kenny ( 1983) on public speaking.

Formal lectures
My intention in writing this chapter (coming, as it were, from the body
of the audience) was to appeal to those giving short talks. A graph of
listeners' comprehension against time shows a fall within 10 min. So
points about good speech and sensible delivery for short talks are
relevant to 50-min lectures too. A joke at 10-min intervals stimulates us
wondrously.

Directive for lulling an audience to sleep

Wear a dark suit and conventional tie; turn down the lights;
close the curtains (drapes); display a crowded slide and leave
it in place; stand still; read your paper without looking up;
read sleadily wilh no marked changes in cadence; show no
pictures; use grandiloquent words and tong scnccnccs..
+o CO MM UN I C ATI NG I N SCIENCE

Appendix on projectors
We are told not to switch projectors off and
on needlessly because that harms the lamp. Switching need not be bad
if a thermistor has been fitted ; every projector should have one. The
projector should carry a spare lamp , in working order.
Overhead projectors have seve ral merits in comparison with slide
projectors, including the following. The room need not be darkened:
this facilitates note taking and the lectern needs no light. Projection is
under the speaker's control. The speaker can see the audience. A
display can be unmasked gradually. Displays can be developed during a
talk. Transparencies prepared beforehand can be written on during a
talk. T ransparencies do not bulge an d go out of focus. Insertion
upside-down is obviated. Transparencies can be prepared, or modified,
without photography and therefore immediately before a talk. Colours
are easily included. One transparency can be superimposed on another.
Big transparencies are easy for interested parties to examine after a
talk. If speakers prepare their own transparencies and write in large
lettering, they are less likely to put on too much than if someone else
makes the slides from typewritten copy.
Against this baker's dozen merits of overhead projectors there are
some demerits, but most of them could be overcome by better design
of the projector and of its immediate environment. Often the overhead
projector is placed on a table or tro lley that is too low and the speaker
has to bend over, which is bad on several counts; also the speaker may
be dazzled when straightening up. If you are able to do so, ask to see
the projector before your talk and to have it raised. I have done that.
Sometimes I wonder, has that trolley's designer [a short person?) or the
head of the institute ever used the projector? A speaker needs more
than a narrow sloping lectern. A flat-topped bench is needed to provide
space for transparenci~s and other items. In some lecture rooms the
projector is sited out of reach of the speaker; much of the merit of the
projector is then lost. A quiet fan is desirable. A curved mirror might
be designed to lessen trapezoidal aberration.
If instruction on the siting and use of an overhead projector were
good, perhaps fewer people would resist using this admirable mach ine.
If you still think slides and a slide projector are better, list their
advantages. Can you produce even a dozen?
Chapter Ex
Empty numbers

Par adox
Scientists make observations to perhaps three or four digits, and process
the results on an eight-digit calculator. Careful scientists then discard
meaningless digits. They also use numbers and numerical words with
care. But some writers use words with a precision corresponding to o nly
one significant digit! Here are examples of these and other aberrations.

Doubtful ratios
The phrase '3 times more than' means '4 times as much as'. If this
surprises you , extrapolate down to once more than , which obviously
means twice as much as. Many readers take 'times more than' to mean
the same as 'times as many as', so the former expression is best
avoided. What does '4 times less than' mean? ' Once more than' means
'twice as much as', so 'once less than' must mean 'none'. Hence '4
times less than' seems to mean 'minus 3 times'.
'Every second baby is a boy' is not true. 'One in every two babies is a
boy' is no better. By omitting 'every' we remove the absurdity. The
words 'on average' might be added. 'Three in every 10 plants were
diseased' implies a regularity that seem unlikely. A fine example (yes, it
is genuine] that shows the absurdity is '81 people out of every 10 ... '.
One's mind tries to visualize 1of a person instead of paying attention to
the statistics. Better would be '17 ou t of 20' or 'more than 8 out o f 10'.
'Decimate' originally meant kill every tenth man . Wise writers apply
discretion in using the term for the e limination of more than a tenth.

Dilution
The term 1 : 4 is a ratio and is read as 'one to four'. For example, ' the
ratio was 1 sheep to 4 goats' means 1 sheep in 5 goats , not 1 in 4. The
41
 42 COMMUNICATING IN SCIENCE

meaning of 1: 4 is different from that of 1/4, which has the sense of!.
Some people read 1: 4 as 1in4 and write 'diluted 1: 4' when they really
mean that the final volume was 4. The '1: 4' is then wrong. Are you
unconvinced? Then consider the term 1: 1, which cannot mean 1 in 1. If
you made 1 vol of solution up to 4 vols by adding diluent, then for
clarity's sake write that. 'Diluted 1 in 4' or even '1-+ 4' may be suitable
for most purposes. If the exact dilution is important, you might write 'l
vol solution A was diluted to 4 vols by adding solution B' . Bear in mind
that 1 vol fluid A (e.g. ethanol or H 2S04) added to 1 vol fluid B (e.g.
water) may not produce 2 vols of mixture.

(I) Are all your numbers correct?

If 'adviser' is misspelt , a subeditor can change that. If27 appears for 21
a context may not help, so we are confused. Editors' time is wasted
over wrong numbers actually detected, but the total of errors is
unknown. In a table, nobody but the author can check the values.
Draw a bar through each hand-written+ to distinguish it from other
characters.

Statistics
Standard deviation and standard error have often been confused. Yet
they are the very quantities that are supposed to indicate precision! So
take care that there can be no doubt and do not omit the M from s.E.M.
When n is small, these statistics have only little meaning. If an author
offers statistics based on n = 2, the message conveyed to readers is that
of desperation rather than useful discovery, and we may doubt the
claims!
An often-seen phrase is 'most probably', which means highest
possible likelihood, i.e. P = 1. Frequent use of 'most' debases its
.
meaning.

Phrases to avoid
'Cages were built between 1988 and 1989.' Quick work indeed, for the
interval is infinitely small.
'Twice the size' is sometimes written for twice the dimension.
Because of possible misunderstanding, the former term should be used
warily or not at all. If a picture's dimensions are doubled, its area is
quadrupled.
 EMPTY NUMBERS 43

A surface 3 m square has an area of 9 m 2 • Some readers may not

know that. Write 3 m x 3 m.
Do not write 3-400 if you mean 300-400.
'A plate 75 x 90 mm' is not scientific. Write 75 mm x 90 mm.
An experiment repeated 3 times is done 4 times. If you have doubts,
consider 'repeated once'. 'Repeated again' might be considered tauto-
logical.
'Divided by a third' is best avoided. If the reader is not interested in
decoding, the information is lost.
'A certain amount' usually means an uncertain or unspecified
amount and is too vague for scientific reporting.
A speaker described a humming bird that lives in high mountains in
Equador. He told us 'To conserve energy the bird's temperature falls to
half at night'. What is half a temperature?
'Nearer 40 than 30' tells us only little: any number > 40 is nearer 40
than 30. We know what the speaker means, but should the language
not have been more precise?

Percentages
The sign °/o indicates a pure fraction without units. The present
tendency of using terms other than percentage for concentrations is
admirable. A writer who describes a concentration as 5°/o and then
writes 'it rose 2o/o' leaves you to guess whether 2 percentage units or
x 1.02 is meant. Percentages are best reserved for comparisons. They
must be clear. Replace '150°/o more than' by 2t times. What does
'250o/o lower than' mean?
Do not write 'g/L of KCI' if the solution is not per litre of KCI but of
KCI per litre. Equally unscientific, though common, is the note that
2 mg/kg of a drug were given. Write '2 mg of drug per kg' or 'drug
(2 mg/kg)'. The argument also applies to 'p.p.m. C0 2'.

Do your readers demand too much?

Perhaps you think these comments are pedantic. Some may be. But
scientific reports, written or spoken, should reflect the precision of the
experiments. Numbers and values submitted to editors or presented at
meetings are sometimes hardly better than 'umpteen zillion' or 'every
so often'.
 Chapter Four
Preparations of the script and figures

The script may be read by several people, usually by an editor, two

referees (reviewers in the USA) , a subeditor (copy-editor) o r language
corrector , the copy-preparer, the printer and the proof reader. This
should be sufficient reason for the preparation of clear script. Yet
poorly prepared scripts are commonly submitted to editors; I have seen
some that were barely legible. Even a script that looks neat may not be
good from the printer's point of view. This is partly because typists'
typography may differ from that of printers as to spacing and punctu-
(1) ation. (Unfortunately the typist may have had to type in that manner to
pass her exams.] After the subeditor and the copy-preparer have
corrected a nd marked the typescript it may almost resemble a manu-
script!
If you prepare the script yourself, please read this chapter carefully.
If someone else prepares the script, please show them this ch apter.

Spacing is important
The script should have wide margins, especially at the bottom. Stan-
dard left margins, say 40 mm wide, will help to ensure that pasted cor-
rections are in alignment and that unintended indentations are avoided.
The script should be double-spaced throughout, including that of
footnotes, notes to tables and list of References. That directive in italics
is something often overlooked. It is not editorial pomp; the space is
needed for instructions to the keyboard operator, and those
instructions are usually more numerous in notes and legends than in the
text . If you still desire to use close line spacing for parts of the script,
make all spelling, punctuation, units, hyphens, capitals or lower case,
decimals, formulas, numbers, etc. in accord with the style of the
journal. If you cannot ensure all that, then leave interlinear space for
the subeditor and copy-preparer.
 PREPARATION OF THE SCRIPT AND FIGURES 4S

The first line of each paragraph should be indented; three spaces are
enough . A paragraph that is not indented (for reasons of fashion?)
risks being ' run on', especially one that starts a page. This abominable
non-indent fashion wastes the time of editors, who have to mark each
new paragraph (NP) and who may have to turn back a page before they
can decide whether an NP is desired. If you dislike indention in a
script, or if the typing is done in a distant office outside your control,
help the editorial staff by marking each NP with a 0 on the final script.
Remember to change ug to µg and ul to µI. Many u's escape
conversion. Ensure that symbols in legends, figures and text agree, for
example 3A or 3a. Make fractions unmistakable; a reader took time to
realize that 2.1/2 was not 1.05 but 2t.
All pages must be numbered, including the first page as well as the
list of References, etc. Imagine what happens if the script is d ropped.
The journal may request two or more copies, and it should be
axiomatic that good copies be submitted. Blurred script strains the
referee's eyes because of the continual attempt- abortively - to
sharpen the image by focusing. If you use a typewriter and its 'e' is full
of fluff, clean it, for example with a toothbrush. Before you make
photocopies, cut a tiny corner off each original page. Then you can
identify the original and avoid making copies of copies. If your script is
printed by a dot-matrix printer, remember to use the near letter quality
mode. A script with fewer than 35 dots per character is unsuitable for
scientific reports.
Typesetters appreciate good copy.

Directives from a Dutch uncle

Do you think these directives are stern? Experience shows that they
need to be. I cannot recall having seen a script that needed no
correction. Some scripts seem not to have been read, in final form , by
the author. If a script has several authors, all of them should read it.
Bad script may not lead to a paper's rejection; but neither will it
encourage the referee to read it promptly. Day [8 (5)) writes that a
badly typed script is immediately returned to the author. Another
editor told me that poor script may delay publication by a month .

Word division at line ends

Do not split a word at the end of a line. If a hyphen appears there, the pr-
inter will usually join the parts of a word. If a hyphen is needed (as
46 COMMUNICATING IN SCIENCE

in dansyl-lysine), use a double hyphen(=) or repeat the hyphen at the

start of the next line. Even better, take the whole word over, because
broken wor- ds interfere with reading. The lines in a typescript need
not all have the s- ame length . Do Jhe unequal lines of this text disturb
you as much as divided words would? When a scientific paper is to be
typed, th- e machine should be instructed not to divide words.

Balloons (p. xii)

In the margin , write legibly or type 'Table .. near here' or 'Fig . .. near
here'. Draw a ring or box round these and other directions to the
printer about Greek symbols, mathematics, etc. Then the directions
will not be printed . Such balloons are useful in other ways too.

Word processor (WP)

When a WP is used, some of the suggestions in this chapter do.not
apply. Corrections are so easily made on a WP that dirty pages should
be corrected and printed afresh.
The WP has such advantages over the typewriter that the latter is
seldom used for scientific script.
On the final copy of your processed script , you may have to add
items manually: special symbols, complex underlinings, instructions,
accents .... If the editor requires 3 copies, you need to make 5,
including one to retain and one to deposit at home. To add a ll the items
to all copies may be a laborious task and some items could be missed.
Photocopies obviate the risk.

Does the VDU cause eyestrain?

Evidence for eyestrain caused by a fluorescent visual display unit
(VDU) is scarce and conflicting.
If you think your VDU tires your eyes, consider the following.
ls your seating suitable?
Do you wear bifocal spectacles? The upper lens is designed for
distant vision , the lower for close work (30 or 40 cm). Your VDU may
be well over 40 cm distant, so neither of you r lenses is ideal. Ask an
optician about special spectacles or a pair with continuously variable
lenses.
 PREPARATION OF THE SCRIPT AND FIGURES 47

Cover sheet
(2) Some publishers appreciate your supplying a top sheet. On it type the
name of the journal , title of the paper, author(s), address for corres-
pondence, running title, number of figures and number of tables.

Corrections and additions to the final copy

After a page has been retyped, check that all has been copied. If I
cannot understand a passage in a typescript, I wonder whether a phrase
has been omitted. Jumps occur when a typist 'picks up' a word in error
at its repeat. (See 'tape ... tape ... tape' five sentences below.)
When a WP is used , sometimes a line is lost or redundant material left
.
1n.
For small corrections, correct the text itself; do not make corrections
in the margin as you would on proofs. Confirm by using a balloon if the
mark is not distinct.
Corrections should not be attached with pins, clips or staples; they
embarrass the printer. If you use transparent tape let it be invisible
mending tape upon which one can write. Common self-adhesive tape
should not be used. The latter, convenient though it is for authors and
secretaries, is disliked by editors and typesetters, comes unstuck and
cannot be written on. Flags, tails, flyleaves, turned-up pages and
additions on the back of a page cause trouble, and their contents may
even be accidentally overlooked by a referee or the typesetter. If more
than a line is to be added, retype the page. Do not crowd the page. If
there is too much for one page, type two pages. Then re-number the
pages through to the end. Although the paper for all pages should be of
uniform size, the lengths of scripts on pages need not be uniform. Short
pages (script, not paper) are not troublesome, but long pages are.
Most of the directives in the paragraph above need not apply where a
WP is used, but they need to be explained so long as some authors
continue to provide typewritten scripts.

The typewriter's or word processor's type face

Before you buy a typewriter or WP, look for a type style that is suitable
for scientific work. Sanserif (p. xvi) is unsuitable. If you have to accept
sanserif and you show 'I' as part of a formula (e.g. In, HI , -ol-1-one),
name the symbol in the margin: ell, eye, ell, one. One day, while you
48 COMMUNICATING IN SCIENCE

wai~ in a queue (US 'line'), you may amuse yourself by recalling

possible meanings of Iii; there are 4 or more. Where there is any doubt
whether o, 0 or 0 is intended, whatever the type face, tell the printer
what the symbol means: oh, zero, cap Oh.

Books on the preparation of printers' copy

Editors and others who prepare scripts for the printer will find useful
information in Copy-editing by J. Butcher and in BS 5261. The Oxford
Dictionary for Writers & Editors is especially useful to editors and
subeditors. These books are described in Chapter Eight (13) onwards.

Numbering of figures, tables and references during their preparation

If you number the tables and figures from first writing you may have to
change the numbers as your paper develops. Instead of numbers, use
letters that describe tables and figures to you privately . Before the final
typing, change to numbers. With this scheme you are unlikely to leave
a wrong number In the text. On a WP, the letters or numbers are easily
found if you follow each with @ or other sign that you otherwise never
use.
Literature cited poses an analogous problem. At first I use the
Harvard System: authors' names and year in the text. If the journal
requires numbers, I change the script at a later stage .

Drawing the diagrams for reproduction

Draw diagrams about twice the dimensions of the desired printed
version. Such enlargement demands thick lines, perhaps thicker than
seem necessary to your eye . The lines of curves should be thicker than
those of axes. Ensure that lines in different diagrams will appear
equally thick, for example by using the same scale for all drawings.
Insert the 'points' of a graph in ink. Sketch the curve first in pencil,
free-hand or against a flexible rule. You can then erase the curve for
re-drawing without losing the points. Or you may be able to calculate
the line of best fit. The Biochemical Society's instructions (8 (2)] give
detailed directions on drawing graphs.
The printer likes drawings to be on smooth card, Bristol board or
good tracing paper, but graph paper is usually acceptable provided its
lines are faint blue. Graph paper helps one to keep drawings of
apparatus square, so diagrams may be drawn on such paper. To
 PREPARATION OF THE SCRIPT AND FIGURES 49

transfer a diagram to Bristol board, fix the former on the latter, then
prick through principal points with a needle. The tiny holes provide a
template for your drawing. They will be inked over, and should cause
no trouble.
Line drawings are reproduced by an all-or-none process. So draw
your lines densely black; faint blue lines, pencillings, etc. are unlikely
to appear on the printing plate. Deletions can be 'made with adhesive
paper. You can make a continuous curve dotted or dashed by sticking
narrow strips of paper over it at suitable intervals. Printed-pattern
paper can be cut to shape and stuck over particular areas to make them
hatched (shaded).
Write words or numerals on a transparent detachable overlay. Many
printers can insert such lettering (labelling) onto diagrams professional-
ly, but this possibility should be checked with the editor. An inexper-
ienced draughtsman has difficulty in making the letters of suitable ~ize.
Examine a journal with author-drawn letters and you will see that
unsuitable labelling spoils a picture. If you draw the letters or apply
(3) pressure-sensitive transfers, use lower-case letters; they are more
legible than capitals. [See 'Legibility of print', 8 (1).] Label each curve
if possible (seeking explanations in the legend is tiresome for the
reader) but do not 'clutter' the diagram. The labelling must be brief, so
as not to dominate the graphs. If there is not room for words on each
curve, perhaps there are too many curves in the figure. Hers (1984)
recommends the use of open ('empty') symbols for control curves; he
also makes other good suggestions.
Draw a magnifier bar with scale value on the picture; then, if the
picture is reduced, the bar is reduced too. A magnification number in
the legend is unsatisfactory on several counts.

Journey's end for the script

If you write many papers or a book, visit a typesetter. You may then
see what can be done and what is typographically costly. If you are
shown examples of good and bad copy, you will see that the above
homily on the preparation of a script is not too detailed. Further, you
should see why alterations at the proof stage are expensive and may
delay publication.
Chapter Five
Addressed to those for whom English
is a foreign language

Errors of particular kinds occur in scripts and talks prepared by those

whose first language is not English. The errors make the difference
between foreign 'engelish' and idiomatic English. Errors of idiom
distract a reader, and errors of pronunciation distract a listener. You
want the reader to be attracted, not distracted. This chapter is intended
to help you.

Words
Certain errors often occur in translation. Sometimes this is because an
English word (control, eventual, sensible) resembles a foreign word
that has another meaning. Other troublesome words are discussed
below.
abolishment is not a word. Write abolition.
acknowledge. One may acknowledge receiving a gift, but one does
not acknowledge a person for help; one thanks him.
adoption is probably not what you mean; try adaptation.
also commonly occurs in a wrong place; the best place is usually
before the main verb. The same is true for already. The editor may
delete the words beca~se they are not always needed in the English
translation.
both. ' Both wires were not hot' leaves it uncertain whether one of
them was hot. Write 'neither wire was hot' , if both were cold.
control, the verb, does not mean count, examine, inspect or observe,
at least not in scientific English. Control means govern, maintain or
limit a variable such as a rate of flow or a temperature. Measure, check
or monitor may be what the writer means. Control, the noun, is well
understood and needs no explanation.
could and able occur too often. 'We could measure' or 'were able to
measure' tells only of the ability to measure. If you actually measured,
so
 TO THOSE FOR WHOM ENGLISH IS A FOREIGN LANGUAGE 51

say so: 'we measured' is definite. In 'the friction could be too great'
perhaps the writer meant ' . . . may have been' .
demonstrate is often written where another word would be better.
The word has various meanings, the most usual being physically to
display or show something happening. 'Nob reports that leaves of some
trees contain gold' is more cautious than 'Nob has demonstrated .. .' .
Show is less grandiloquent than demonstrate, but is not always suitable.
Write 'the results show' (not demonstrate); but write 'Cob has found'
rather than 'Cob has shown' because he was writing about his experi-
ments not showing them. 'Table 3 shows' is better than 'It can be seen
in Table 3'. Usually, 'it has been demonstrated that' may be omitted: if
you write about another's observation, write in the present tense; the
meaning should then be understood and the five words are not needed.
If you saw red grains ~mong predominantly green grains, write that
they were seen, not demonstrated.
describe is a transitive verb. You may describe a method or an
apparatus. But 'Blob has described that chopped straw makes good
fodder' is not good English. Write that he claims, states, writes or
reports, or has claimed . ...
dosls is not in common use; write dose.
eventually does not mean maybe. The event will happen - ultimately.
experience should not be written for experiment.
filtrated should be filtered.
insignificant, which means unimportant, should not be used in
statistics. Write 'were not statistically significant' and quote the P
value, or write non-significant. If you have no P value, you may write
negligible.
know should not be used in the sense of to provide knowledge. 'The
fossils were studied to know . .. 'is not good. Write 'we studied the
fossils to find out . . .'.
obtain means acquire, be given. Manipulations provide (not obtain)
material. Obtain has the sense of acquire or fetch rather than of give.
quantitate is not (yet) a word in English dictionaries. You may
quantify the effects though that is a grandiloquent word for measure or
count.
registrated is not a word. A variable may be registered on an
instrument such as an altimeter; readings from it are usuaUy recorded
rather than registered.
resorption should not be written for absorption, except in re-absorp-
tion, or for the special case of resorption of a foetus (fetus), limb or
other part originally produced by the body.
S2 COMMUNICATING IN SCIENCE

respectively. In 'we cleaned the sherds and tiles, respectively' the

word respectively is not needed, but is needed in 'we cleaned the
sherds and tiles with water and oil, respectively'. Change 'blue
respectively red' to 'blue and red, respectively' , and do not shorten to
resp.
since. Write 'seven years ago' not 'since seven years'.
supposed may be bettered by 'believed' , for example , in 'birds are
believed to have evolved from dinosaurs'.

Idiomatic English
Some common errors of idiom will now be discussed.
'Powder was added and stirred' is not good sense . Write 'Powder was
added and the mixture was stirred'.
The abbreviation a.o. is not in common use, but 'etc. ' is. The
abbreviation for number is no. not nr. [1 (13).] The term aqua dest. is
best avoided.
'This permits to do that' needs an operator after permits; e.g. us.
Patient, the noun, is a sick person ; patient, the adjective, means
having patience. So patient blood means blood having patience! Write
'blood from patients' or 'patients' blood'.
Things compared must be comparable. 'Growth was similar to the
controls' needs 'that in' before 'the' . 'Resonances in pipes were unlike
rods' needs 'those in' before 'rods' . 'Lymphocytes from treated patients
were larger than untreated patients' needs 'those from' after 'than'.

Spelling and pronunciation

Certain words are commonly misspelt : occurred, subtract, oxidation,
centrifuged, homogeneous, synthesize, desiccate, naphthol, phthalic,
saccparide. If the pronunciation of a word is changed , the meaning may
also change. Lead, the noun , pronounced ledd, means Pb ; the verb,
pronounced leed, means to conduct . Live, pronounced !iv, is a verb
that means to be alive; pronounced lyve, it is an adjective describing an
organism that is alive. Never pronounce the word as leave , which
means go away. 'Live [lyve] fishes may be seen in the river where they
live [!iv]'. [Do you think English is illogical? Yes, it is.]

Use of a dictionary to check meanings of English words

In your Spandanese- Engels dictionary, you may find a word you want
in English. Next , back-check the English word in your Engels--
 TO THOSE FOR WHOM ENGLISH I S A FOREIGN LANGUAGE S3
Spandanese dictionary. Then consult an English dictionary (Chapte r
Eight) to confirm that the word is what you want. This procedure will
help you to avoid writing spatial cubicle when you mean spacious
cubicle, vaulted caterpillar when you mean arched caterpillar, favourite
for favourable, sensible for sensitive, the colour is exactly different, or
that the animals' diet was spiked with vitamins. You might be
embarrassed were you to see, in print, that you had written permis-
siveness for permittivity.
While you are using the English dictionary to check the word, note
how it is pronounced.

Speaking at conferences
Unless you have often conversed in English, try to find someone who
will check your pronunciation. If you empha'size incorr'ect sylla'bles
the audience has to interpret words and may lose a sentence during the
time it takes you to do that. A lost sentence may jeopardize your
argument. So just one mispronunciation may ruin your talk. Forestall
such a tragedy by displaying important words in writing.
I tried to follow a talk where the key word was spoken many times,
always mispronounced, but never written down. For me, at least, the
message failed.

International conventions
Use the international decimal point (0.6 g) for papers in English unless
your publisher prefers the comma (0,6 g). Write 100000 not 100.000 or
100,000 if you mean one hundred thousand.

Encouragement
For your comfort I may add that English people, too, find it difficult to
write good English.
When you visit England or any other English-speaking country, or
when you watch and listen to television from such a country, you may
be appalled by the poor quality of parts of the speech you hear. You
probably have a better knowledge of grammar than do the inhabitants.
You may even be able to help to maintain the quality of what is now
the International Language, especially when you become an editor. I
hope you will do that; I have known Continental European scientists
who have done so and I have learned from them.
Chapter Six
An appeal to North Americans

There are more of you than there are of us British. So you are now the
Trustees of English, the International Language. Sad to say, not all of
you are taking your trusteeship as seriously as you might.

One language
English , at its best, is much the same on both sides of the Atlantic. We
spell some words differently, but we mostly understand one another.
Some words are spelt more phonetically in the USA than in Britain .
One day perhaps we shall have a unified spelling. Several words for
food , transport and domestic items are different; but in science the
differences are few.
Please understand, I do not suggest that, where American English
and British English differ, the British version is always the better. The
form more easily understandable by T. W·. Fline is the better. T. W.
Fline are Those Whose First Language Is Not English. Imagine them to
be looking over your shoulder as you write.

Grammar
British and American writers may be equally remiss in their treatment
of English grammar, so I refer you to the section in Chapter One
(p. 8).

Complex clumsy long-winded adjectival phrases

What Woodford (8 (3)) calls stacked modifiers are prolix and at times
barely comprehensible noun-adjective and modifier groups, as are this
one and the heading above.
Consider 'voluntary human kidney donor research institution person-
s+
 AN APPEAL TO NORTH AMERICANS SS
nel'. A reader sees 6 words before realizing that the writer means
people not kidneys. The reader of 'barley root tip cell chromosome
aberrations' sees 5 false nouns before finding the real noun. The five
checks or hold-ups tire the brain. The phrase would be better as
'chromosomal aberrations in root-tip cells of barley'.
The US journal Science carried correspondence on 'adjective noun
use tendency'. Hildebrand (1983) gave examples of shockingly cum-
bersome phrases, and suggested that 'of' should be used to improve
them. On the other hand, Baer (1983) disliked 'measurement of the
angle of the joint of the ankle' and thought 'ankle-joint angle measure-
ment' was better. Better still would have been 'measurement of the
ankle-joint angle'.
So far as practicable, stacked modifiers should be avoided in writing
and in speech. (See 1 (7).)

Future tense
Students in non-English-speaking countries learn that shall and will are
used formally for the future tense but that to be going to is usual in
speech. Which do you prefer: 'he's going to repeat' or he will repeat?

Americanizations
An Americanism that makes editors wince is the making of long words
by adding -ize, -ism or -ization. Examples are prioritized, de-
logarithmization and summarization. If you see such words, perhaps
they will remind you to persuade colleagues and your research students
to abstain from such 'manufacturization'.

Short words are often preferable to long words

T. W. Fline should not be conditioned, by example, to use ponderous
words where simple words are adequate. Here are examples: method-
ology rarely betters method; enzymatic is less elegant than enzymic;
detoxification is no improvement on detoxication; are experimental
animals killed, or sacrificed? Subsequently to is a grandiloquent
synonym for after. Following is no better, as this example shows: 'A
still-thirsty cat was seen following a drink.'
In most US journals, the 'al' has been dropped from physiological,
symmetrical, serological and other words. If these shorter words are
acceptable, can you drop other needless syllables too? for example
from sonification.
56 COMMUNICATING IN SCIENCE

Export
You may point out that complex modifiers and ill use of words occur in
European and other writings. True, but (a) not so often as in the USA,
(b) they may have originated in the USA, and ( c) that is no defense for
their perpetuation in US jo.umals or books that will be sold abroad.
Please read again the first words of this chapter.
When Americans export words they should be careful to ensure that
what they export is good. The export may be unwitting, but it occurs
none the less, through journals, conferences, television . . . . People
abroad copy what Americans write and say - the bad as well as the
good. 'Prior to' , for example, is becoming common in foreign writing;
'effect' and 'affect' are wrongly interchanged, as are alternate and
alternative.
Please take care over these and other words, including such vogue
words as basically (usually best omitted) , dimension, situation, exotic,
framework and secret.
In chromatography, a column of adsorbent, sometimes called the
bed, is held in a tube. The tube is the support; it is not the column. In
the USA, the word column is often used for either. This confusion has
been exported; so, when a writer in any country states the length of a
column, the reader cannot always tell whether the length is that of the
container or the contents; the latter is what matters.
Corn , a general term abroad, means grain or cereal. In the
Americas, corn usually means Indian corn or maize ; in wheat-growing
countries, corn may mean wheat. So, when a scientist speaks or writes
about maize, or maize oil used in diets, he or she should give its Latin
name (Zea mays) at the first mention, because some of the listeners or
readers may be foreign. Likewise, 'corn oil' should be defined.

Homonyms, words with more than one meaning

Homonyms are a fact of everyday life; but in scientific writing, where
communication must be precise, certain homonyms can introduce
ambiguity. This problem is multipled in some transatlantic transfers.
For example, formerly, to fix meant to make secure; in the USA, 'fix' is
used for dozens of meanings including mend, destroy, mix, arrange and
repair. This plethora of meanings confuses foreigners whose lexicon
tells them that fix means fasten. Does 'he fixed the blockage in the
apparatus' mean that he made it permanent or that he cleared, i.e.
unfixed, it? In mutation research, 'genetic lesions were fixed' has the
 AN APPEAL TO NO R T H AMERI CANS 57
opposite meaning of genetic lesions were repaired. When 'repair' is
eventually Jost, perhaps the word will persist only as jargon in
molecular biology!
Most of the words in the USA that have so far taken over the
meanings of other words are used generally rather than for scie nce.
However, unless the Trustees are vigilant, this 'homonymization' will
continue to invade scientific communication too, as the following
examples show.
Localized is sometimes used for located. The latter means that
something is situated, or was found , there; the former means it is only
there.
Thus is often used where hence, so, there/ore or evidently might be
better. The original meaning o f thus was in that way. Should we not
keep it thus?
Manuscript means something handwritten. Neither a typescript nor
the product of a word processor is a manuscript. So let us leave the
word to describe manuscripts, and use 'script' for a mechanical
product. Thus (=in that way) we can make a distinction and ofte n save
4 letters.
For transportation, i.e. the act o f transporting, transport (train, cart,
bus etc.) is used for the purpose . In the USA the longer word is used
for both meanings. On occasio n , the two meanings might be needed
unequivocally in the same passage.
Practical and practicable are sometimes inte rchanged. The wearing of
goggles in a laboratory is practicable in the sense that one is able to
wear them, but vizors are more practical because they are ventilated.
A referee reads a paper and m akes recommendations to the editor.
A reviewer writes reviews for publication. There are two words fo r two
meanings. In the USA , reviewer is used for both meanings.
Diet formerly meant the food of an animal, including humans. Now
diet has become common usage for restricted food for certain people.
Sometimes we need the early mean.ing, for example in studies on
energy values. (See next paragraph.] We need another word.
Animals are given a diet, no t fed the diet. One may feed an animal,
but one cannot feed a diet. A US author 'fed a piscivorous diet'. He
probably meant that his animals, not his diets, were fish eaters.
If bred is written for mated, readers outside the USA may not
understand. Mated is understood internationally to mean that a male
and a female were brought together for reproduction. Bred has another
meaning: after many selections and matings [or cross-pollinations) a
new strain of a nimal [or plant] may be 'bred'.
S8 COMMUNICATING IN SCIENCE

In Europe, stones and rocks have different meanings. By using rocks

for both, you may be losing a useful distinction. If you say rocks, a
European might not know that you mean stones, rocks being thought of
as bigger than stones and usually fixed to the world.
Ensure and insure have two meanings, two spellings and two
pronunciations. Only 'ensure' (to make sure) is likely to be needed in
scientific communication, but do you not think we should keep the
separate spellings?
Practice (noun) and practise (verb) are two words abroad. Has
something perhaps been lost by the confusion of the spellings in the
USA? A similar question might be asked about licence and license or
tyre and tire.
The adjective a/ter'nate refers to a happening to one thing then the
other; alter'native implies a choice of one or the other. To al'temate is
the verb. Electric current that alternates is called AC; DC is an
alternative to AC. In an AC circuit, each pole is alternately positive
then negative and so on.
Substantive is not a good alternative for substantial, because nowa-
days the former is used in specific legal and political phrases.
To quit means to go away from, to leave. In the USA the word is
used to mean cease. In scientific writing it would be better to use stop,
desist, refrain . . ..
With words being 'sacrificed' or made to double up, T. W. Fline
hope their Spandanese-Engels lexicon will not become out of date
faster than the publisher can cope with change. In a US dictionary, you
may find some of these alternative meanings of given words. Diction-
aries do not instruct, but report usage. For T. W . Fline's sake, please
do not provide them with more material for 'homonymization'.

Careful choice of words

An essential of good English is that each phrase shall have only one
meaning. That cannot be achieved if words 'stand in' for others. A
scientist who wants to be understood abroad should remember that
clarity is more important than whimsical fashion.
Hopefully is often used ambiguously. Were I to write 'Hopefully you
will agree with some statements' that would mean you (not I) hope. So
it would be truer to write .'I hope you ... '.
Due to and owing to have different meanings , as explained at I (12).
Maybe it is too late to save pistol , stones, mend, locate, owing to ,
 AN APPEAL TO NORTH AMERICANS 59

candy and other words with particular meanings. However, if the decay
continues, your grandchildren may not be able to read and understand
Mark Twain's books. Does that sadden you?
Of course you cannot be expected to give up established idioms, such
as 'this moment in time', face 'up to' a problem, meet 'up with' a
colleague, institutionalized, flat, make (for arrive at). Nevertheless,
examples are given in this chapter to remind you of the trend and to
reinforce the Appeal to you not to let other confusing usages or lengthy
phrases become established. A professor in Minnesota said you cannot
reverse the flow of the Mississippi; however, in view of your other
engineering feats, I believe you could control the flow.

Speaking abroad
Naturally, you may not want to see the demise of local speech, be it
Southern USA, Scots, Cockney or other. Each of us may think his or
her way of speaking is correct and that the world should concur. The
world does not concur: pronunciation depends on usage. Words are
spoken differently in the Americas, Europe and elsewhere, so care
must be taken over pronunciation whenever some listeners may be
from outside a speaker's region .
The southern English pronunciation of path as parth seems odd to
Americans. Likewise the American softening of 't' that produces
twenny, madder and wahder puzzles foreigners. Metal mercury and
methyl mercury are poisons that act differently. In a television program
a speaker pronounced metal as meddel and methyl as methel (soft 'th'
as in weather). At least one listener outside the USA could not
distinguish the two words, so the gist was lost. In a lecture, a foreigner
was baffled because 'isolated' was pronounced issolated. A Californian
professor pronounced carotene as keratin. This caused a distraction
until listeners became used to it.
Whether you say ameeno acid or amyno acid is unimportant.
Whether you say vytamin or vittermin, we shall understand you.
However, the formerly mysterious 'accessory food substances' were
known to be vital and were at one time believed to be amines. So 'vital
amines' became vitamins and were pronounced vytamins.
When you take part in an international congress or in a television
program, speak slowly. If you have so much to tell that you must race,
you will not achieve your objective if listeners cannot understand you.
To gain time, use short words ('use' (noun) not 'utilization') and utter
60 COMMUNICATING IN SCIENCE

no needless words (very, really, invariably, up to). If we cannot

understand you, we shall not respect you; but if we enjoy hearing (or
reading) what you tell, we shall revere you.
One way to help listeners is to speak according to CBS English. I
would find that difficult, so I understand that others might too. At least
it should be tried for technical terms, which need especial care.
A good way to help yourself to speak well is to write a book on
lecturing. Please write it for an international readership.

Papers submitted to foreign journals

Many US writers send papers to European journals. One editor said
the best papers are excellent but the worst are not as good as those
from writers whose first language is not English; he thought the spread
of quality was too wide. My editor at CUP, evidently knowing this,
asked me to write the present Appeal.

.
Caring Americans ARE concerned about the trend
In case you think that I am biased, let me remind you that authors of
American books on writing also dislike the bad features discussed here.
Houp & Pearsall , in their excellent book [8 ( 4)), give a list of
pomposities to be avoided. If you have not read their Chapter 8, you
have missed (eYt eR) a treat.
Day [8 (5)] instructs scientists - firmly- to use short words.
Woodford's (8 (3)) piece on stacked modifiers, mentioned above,
deserves study. He wrote it in the USA.
Nicholson's American-English Usage [8 (9)), though out of print,
deserves to be in print. She argues eloquently against careless writing;
her own prose provides pleasant reading.
Holman (1962) writes ' Inelegant writing may charm the writer, but
... offends the reader'. He quotes several abominations, including
'We horizontalized the patient ... and decholecystectomized him'.
An American scientist (David E. Green, later to become a professor
at the University of Wisconsin) instructed me, a research student, in
scientific writing. He taught me, among other things, only to write an
'it' that would easily be related to a noun. He was careful to say ' I have
to' not 'I gotta'. I learned much from him.
Editors of several US journals are evidently particular about English.
Articles in Scientific American, for example, as well as papers in
various other journals, are (mostly) noticeably well written and edited.
 AN APPEAL TO NORTH AMERICANS 61

Editorials in The New Yorker are written superbly. The New England
Monthly has established the Grammar Police to bring errors to the
attention of offenders.
The annual Gobbledegook Award is an American institution.
Obviously, then, som_e Americans do care, and care very much,
about the future of the language of science. Alas, they are not a
majority. If more US scientists wrote as one of you (Strunk, 8(10))
recommends, this chapter might be unnecessary.

How large is your vocabulary?

An ample vocabulary is a necessity for every scientist, although the
interpretation of ample is arguable.
Will you do an experiment? Open a dictionary; on one page,
count the words whose meanings you know; turn to other pages, then
count ... ; divide your total by the number of pages you examined;
multiply this average by the number of pages in the book. Then
compare your grand total with the three-quarter million or more words
in English. I have done that; the result was humbling. If your score is a
single-digit percentage, consider keeping a dictionary near the dining
table. When the family argues about a word, consult the dictionary.
That will increase the family's word power.

Useful American words

Sad to say, Britons have dropped some early words that you have
kept (gotten, fall, closet ... ). The US use of presently to mean at
present betters the UK use to mean soon. Faucet is hardly known in
Europe, although it betters tap because the latter has various
.
meanmgs.
Nevertheless, the world does know many of your words. 0 Kand
radio are universal. Hindsight, dangling participle (such as 'offspring
were born 6 weeks after mating') and blurb are in common use. Opine
(though many hate it and only few say it) is more apt than figure, which
does not mean believe. The world knows that a wrench is the same as a
spanner, even though the former hints at cruelty to nuts.
American wordsmiths have invented these words. So why do they
not invent others when they are needed instead of using existing words
for new meanings? Alternatively, since Britons accept so many of your
words, why not accept some British words? Or consult Roget's The-
saurus? [8 (17).)
62 COMMUNICATING IN SCIENCE

Words that give no useful information are not needed and could be
left out
Scientists of all nationalities write and say what Houp & Pearsall (8 ( 4) J
call empty words. The use of certain of these words was, until recently,
peculiar to North America. 'Right' is an example; in what way does
'right here' differ from here? The superfluous 'right' and other words
(up with, out on . .. ) are right now invading 'up' Europe and are being
uttered in scientific talks. The Trustees should set a good example by
leaving them out, unless a special shade of meaning is desired.
Because Americans invent such pithy phrases as 'Tuesday through
Friday' and because t= $,is it not odd that they write 'in the
neighborhood of' or 'along the lines of' for 'about', and 'in the event
of' for 'if'?

Data
Data once meant things given (or known) that could be used in
argument. Readings from an instrument were not yet data. The raw
results had to be averaged and perhaps processed in other ways. The
data so derived could then be used to support or refute a hypothesis.
Not only is it too late to reclaim the early·meaning, but a further
change looms, namely singularization. (What a word!) Data was plural.
Many people would keep it so. If you agree, you may also think that
'the data presented' or 'the given data' is tautological.
Data is pronounced variously (dayta, dahta, datta, dadder], so speak
the word clearly. A speaker may say ' Right now we're gunna subject
the data to computerization' . During the time it takes us, the bemused
foreigners , to interpret the phrase to mean 'We shall now analyze the
results', the speaker may utter another sentence which we shall miss.

Evolution of language
Languages evolve; English is a blend of at least three. Slang becomes
respectable. We cannot fix English for ever. But, instead of letting it
slide aimlessly, we (all of us) should pay attention to the manner in
which we let English change.
A professor of linguistics explained that the USA, a 'young' country,
innovates fast and that Americans use an existing word in a novel way
'for the sake of change'. On the other hand, it seems sensible that
useful words should be kept alive, including shall, will, results, maize,
 AN APPEAL TO NORTH AMERICANS 63

alternative, locate. Irregular verbs and irregular plurals could be made

regular. [Agendas and medias are already on the way in.] A new word
might be made for one of the meanings of a word that now has two,
e.g. plasma.
T. W . Fline would welcome such changes, and others too. They
might add 'Why may we not write "two times", sheeps or depend
from?'
Too much rigidity would stultify the language, but, if we must alter
it, let us be sensible in doing so.

Our obligations to T. W. Fline

If you think I am unkindly critical, please bear in mind that many
people hold similar views and that several of the examples have been
provided by editors. We, Americans, British and others, are fortunate
in having the International Language as our mother tongue. Let us help
T. W . Fline by restraining abusage and improving good usage.
Chapter Seven
Preparation of a doctoral dissertation
or thesis

A thesis is like an unusually long paper, includes a review and is usually

divided into chapters, so it resembles a book. Much of the detail that is
described in other chapters of this monograph applies to the writing of
a thesis. Although the style should be concise, no part should be so
brief as to risk being inadequate. (The words concise and brief have
different meanings bu.tare often confused.]

Specialization
If you have worked mainly on your own, rather than in collaboration
with a supervisor, you may have come to know more than anyone else
in the world about one narrow subject. So you must explain your
problem fully.

Collaboration
Most research is done by teams. If you worked with others, you must
do your utmost to make it clear what parts of the work reported were
yours. The examiners (assessors) will be especially curious to know
how much you contributed to the thinking, the initiation and the
conclusions of joint work. Write about this in a Preface.

Your own work

At the oral examination, the examiners, in asking questions about your
interpretations, may be trying to find out whether your thesis was
genuinely your own work. At one extreme the supervisor or advisor
(US) almost writes the thesis. The other extreme was shown by the
attitude of my supervisor: when I asked whether he would comment on
the thesis I had written for a fellowship, he declined on principle and
64
 PREPARATION OF A DOCTORAL DISSERTATION OR THESIS 65
was somewhat shocked at my asking. If you write the thesis entirely
yourself, use the same words as you would in deliberate speech;
grandiloquenrwords will do no good. If English is not your native
language, and someone helps you with language correction, obtain
their permission to give their name in the Acknowledgements.

Imaginative evaluation of previous reports

In a thesis, the literature may be more fully reported than in a paper.
Even a history of your problem may be acceptable if you can make it
interest~ng. Discuss published work critically but not unkindly. Exam-
iners like to know that you can evaluate other people's reports.

Looking ahead
Your inconclusive experiments may be mentioned and used as a basis
for suggestions about what might be done next. Indeed, unlike a
scientific paper, a thesis is a suitable place in which to propose
experiments to test a hypothesis. [But see Promises, 1 (21).] Intelligent
speculations, too, may have a place in a thesis. The examiners may
even ask you about future hopes and intentions.

Allow enough time for careful correction

If you heed the advice given on p. 2 (When to begin writing), you must
start your thesis before you are two-thirds through your course. You
may think this is absurd because you have so few results. Even so, you
can start on certain parts, as suggested on p. 3. The examiners may be
seeking answers to the questions 'Does this student know how to set
about a problem? Can he (or she) think?' rather than 'Has he (or she)
done many experiments?' .

Physical presentation of the script

When you set out a thesis, consider the reader. Do not merely copy the
style of another thesis or offer a cluster of assorted reprints and
photocopies.
Examine well-designed books and arrange your thesis as a printer
would. Write many headings or sub-heads (some people call them
captions] and make a clear distinction between different kinds. Copious
headings in the Discussion are especially important. Pages of script
without sub-heads look stodgy.
66 COMMUNICATING IN SCIENCE

If you have already published papers on subjects related to one

another, combine them into one narrative, then type the whole. ('Type'
includes composing on a WP.] Type the combination yourself even if
you later give it to a professional typist for final preparation. As you
type, you will edit and improve the script more effectively than if you
only read it. If the thesis has many non-standard signs, symbols or
pictograms, it may be well to consider producing the whole thesis
uniformly by calligraphy. First find out whether that is allowable. The
call~graphy must be clear and simple without swash letters.
Attend to the recommendations in Chapter Four. Examiners who
find sloppy presentation may suspect that the experiments were sloppy
too. Leave ample margins, with that at the foot greater than that at the
top. A scanty bottom margin makes the script seem to be sliding off the
paper. If the paper is A4 or near that size and the sheets are stab bound
(stapled through the margin), the script should be no wider than
140 mm; a wide left margin is essential. [This may not apply in
countries where the thesis is printed as a book.) If the bound volume
will be too thick for stab binding, perhaps you should make two
volumes. If, for this or any other reason, your thesis is in two volumes,
make it clear that this is so, both on the covers and on the title pages.
The two assessors of one Ph .D. student each examined a different
volume without realizing there was another.
Give a complete table of Contents (but do not call it an index) and
another of figures. Of course, if you can supply an Index, that would be
impressive, but there is rarely time.
Explain the abbreviations you have used and arrange them as a list
on a separate page or pages.
Place the Summary before the Introduction.
Several small tables are better than a few large ones. Place them, and
the diagrams, near the relevant text; the smaller the table the easier
this becomes. Arrange tables, diagrams and pictures upright (portrait
fashion); readers dislike having to turn books sideways to study
'landscapes'.
Think about response/dose curves and related subjects in logarithmic
terms. If one adds 1, 2, 3, 4, 5, 6 units of reagent, the jump from 1to2
is 1 unit but the concentration is increased x 2. The jump from 5 to 6 is
also 1 unit but the relative increase is only x 1.2. For some studies it is
better to arrange that each rise is proportionally the same. If your
graph is then crowded at one end, try a non-linear scale. Convert doses
into log doses or use log graph paper; or try reciprocals, squares or
square roots; try dissimilar conversions for the two co-ordinates. Such
 PREPARATION OF A DOCTORAL DISSERTATION OR THESIS 67

(I) conversion may use the graphic area more effectively, and it may reveal
a straight line.

The Conclusion
Your Conclusion, if you have one, warrants a section or even a chapter
to itself. If it is only a few lines long, leave it so; do not pad it out.
Raw results must be processed or digested, by you, to provide data
for use in your argument; but numerical values should not be promi-
nent in the Conclusion. The latter should, in general, be in words
supplemented by a minimum of numerical values if the subject allows
that.

You should correct the script yourself

Every research student should read Clark (1960) on preparing for
research, if the book can be found. The prose is a joy to read.
Various books have been published on thesis writing. A good essay is
that by Hawkins (1982). He warns that an assessor may check some of
the references for accuracy. He writes about 'errors that have occur-
red'. The errors include: misquotations; tables upside-down; bad
captions; masses of complicated data with no explanation; statistics that
make no sense; incorrect references; words left out; and literal errors.
The use of the colloquial 'horrendous' seems appropriate here. So be
painstaking over the preparation, reporting of results, spelling and
punctuation; there will be neither editor nor printer to put those in
order for you. The thesis will be with you for life.
One often finds the standard error of the mean confused with the
the stand deviation . So never omit the M. from s.E.M.
Did you find the two errors in the paragraph immediately above? If
not, look again; then let that be an example to show how carefully one
must correct a script.
After you have prepared the first draft of your thesis, read again
pages 7 to 26 herein.
If the examiners find faulty logic, they will tell you, even if it is as
trivial as 'The patient had anaemia because I observed a low r.b.c.
count', or if you imply that the animal body contains serum.

Do not be discouraged
Are you overwhelmed by all the details that need your attention?
I hope not. Attending to details takes time, which may prevent
68 COMMUNICATING IN SCIENCE

your doing all the experiments you would like to do. May I remind you
that
' A good thesis based on few results
betters a bad one based on many'?
Even though they point out errors, the examiners are probably
sympathetic. They will not fail you for a few small faults if they are
convinced that you are a good researcher and that you are able to
communicate and, above all, that you can think.
 Chapter Eight
Further reading

References
Baer, D. M. (1983). Adjectives, nouns, and hyphens. Science 222, 368.
(1) Booth, V. H. (1960). Legibility of print. Research 9, 2-5.
Clark, G. Kitson (1960). Guide for Research Students Working on His-
torical Subjects. Cambridge University Press.
de Bono, E. (1967). The Use of Lateral Thinking. Pelican Books,
Harmondsworth, UK.
Dixon, H. B. F (1983). Return of the dalton. Trends in Biochemical
Science 8 , 49.
Hartree, E. F. (1976). Ethics for authors: a case history of acrosin.
Perspectives in Biology & Medicine 20, 82-92.
Hawkins, C. (1982). Write the MD thesis. In How to Do It, pp. 52-61.
British Medical Association, London. See 8 (7).
Hers, H.-G. (1984). Making science a good read. Nature '307, 205.
Hildebrand, M. (1983). Noun use criticism. Science 221 , 698.
Holman, E. (1962). Concerning more effective medical writing. A plea for
sobriety, accuracy and brevity in medical writing. Journal of the
American Medical Association 181 , 245-7.
Kenny, P. (1983). Public Speaking for Scientists & Engineers. Hilger,
Bristol.
Mackay, A. L. (1977). Harvest of a Quiet Eye. Institute of Physics, Bristol.
Maier, N. R. F. (1933). An aspect of human reasoning. British Journal of
Psychology 24, 144-55.
Norman, P. (1980). Sunday Times Magazine 1980-03-02.
Norris, J . R. (1978). How to give a research talk: notes for inexperienced
lecturers. Biologist 25, 68-74.
Perttunen, J. M. (1975). The English sentence. Luonnon Tutkija 19,
113-17.
Roland, C. G. (1976). Thoughts about medical writing. XXXYII. Verify
your reference. Anesthesia & Analgesia . .. Current Researches 55,
717-18.
Wolner, K. A. (1975). Claude Emile Jean-Baptiste Litre, International
69
 70 COMMUNICATING IN SCIENCE

Newsletter for Chemical Education (International Union of Pure &

Applied Chemistry), No. 11, 7-9.

Instructions to authors
Many journals issue Directives to authors. A good example is the
(2) Biochemical Journal's Instructions to Authors, which includes a piece on
Policy, useful lists of abbreviations symbols, etc. and a reminder to write
RN Aase and DNAase. Obtain the booklet from the Biochemical Society
(59 Portland Place, London WIN 3AJ).
If you have not seen a questionnaire that editors send to referees, try to
obtain one. Then ensure that your paper would elicit satisfactory answers
before you submit it.

Books on scientific writing

(3) Woodford , F. P. (1970). Editor, Scientific Writing for Graduate Students ,
Macmillan, New York & London.
(4) Houp, K. W., & Pearsall, T . E. (1977), in their Reporting Technical
Information, Glencoe Press, New York, and Collier-Macmillan, London,
have an entertaining chapter on writing in clear English.
Scientists Must Write: A Guide to Better Writing for Scientists, Engineers
& Students, by Robert Barrass (1983) , Chapman & Hall , London , includes
a section on reading the literature, which is unusual.
(5) How to Write and Publish a Scientific Paper, by Robert A. Day (1988) ,
ISi Press, Philadelphia, and Cambridge University Press, gives a wealth of
advice on how so to write that your paper will be accepted for publication.
A Manual of Style, University of Chicago Press, is one of the best-known
style books in the USA.
(6) How to Do It (1982), British Medical Association, London, is the most
versatile of these books. Written by medical doctors, the essays tell how to
write, to speak, to chair a conference, to attract a reader, to referee a
paper, to be a dictator, to examine, to be examined, to use a library and to
perform two dozen other activities.
Science Writing for Beginners , A . D. Farr (1988), Blackwell, Oxford.
Better Scientific & Technical Writing, by M. I. Bolsky (1988), Prentice
Hall, New Jersey.
(7) The Chemist's English, by Robert Schoenfeld (1985), VCH Verlags-
gesellschaft, Weinheim. A book of entertaining essays.
Research: How to Plan, Speak & Write about It. Edited by Clifford
Hawkins & Marco Sorgi (1985). Springer-Verlag, Berlin. Eight essays by
various writers.
How to Write & Publish Papers in the Medical Sciences, by E. J. Huth
(1982), ISi Press, Philadelphia, includes a piece on ethics.
 FURTHER READING 71

There are other books on science writing, including a version of this

book in Japanese, To kyo, Jt!! ).. • 8 (Chijin Shokan).

English usage
(8) H . W . Fowler's Modern English Usage, as edited by E . Gowers, Oxford
(9) University Press, or by M. Nicholson (as American-English Usage, Oxford
University Press Inc., New York) , is invaluable. The Nicholson edition is
out-of-print but well worth seeking.
E. Partridge's Usage & Abusage: a Guide to Good English (1982),
Hamilton, London, is preferred to Fowler by some editors.
Some passages in each of these books are slowly becoming outmoded.
Everyman's Good English Guide, by Harry Fieldhouse (1982) , Dent,
London, is more contemporary than the other works, but not so extensive.
The book gives help on many troublesome words and recommends
pronunciation. The grammatical section is easily readable.
Another helpful book is E . S. C. Weiner's (1983) The Oxford Guide to
English Usage , Oxford University Press. This book includes a Glossary of
troublesome words.
John 0 . E. Clark's Word Perfect (1987), Harrap, London, is a dictionary
of current usage. Entries include words that are often confused, errors of
grammar and advice on using cliches, vogue words and foreign words.
(10) Strunk, W. (1959). Thi Ele"1ents of Style (various editions, e.g. with
E. B. White), Macmillan, New York.
(11) What a Word! by A. P. Herbert ( 1935), Methuen, London, who calls
(12) you Bobby, shows that good style need not be dull. B. Dixon (1973) in
Sci write (Chemistry in Britain 9, 70-2) urges lively writing and gives
examples of stuffiness.

Books for editors

(13) Copy-Editing, third edition, by J . Butcher (1992), Cambridge University
Press, contains a great amount of information, some of which cannot easily
be found elsewhere. The book describes the astonishing amount and
variety of work that has to be done on a script and its illustrations between
its receipt by the publisher and touching of the first key by the typesetter.
It covers up-to-date methods of book production , including the 'electronic
typescript' .
(14) Editing Scientific Books & Journals , by Maeve O'Connor (1978),
Pitman, Tunbridge Wells, is excellent; every editor should have this book.
The Oxford Dictionary for Writers & Editors, Clarendon Press, Oxford,
helps with many problems of detail (capitals, spelling, abbreviations,
confusables . .. ) that only specialists can remember in total.
The British Standards Institution (2 Park St, London WlA 2BS) has
 72 COMMUNICATING IN SCIENCE

issued various standards of interest to scientists and editors, e.g. BS5261

(1976) on Copy Marking & Proof Correcting.

Dictionaries
(15) The Concise Oxford Dictionary, Oxford University Press.
Chambers Twentieth Century Dictionary, Chambers, Edinburgh.
The Collins Concise Dictionary ofthe English Language, London.
Reader's Digest Universal Dictionary, London.
The Oxford and the Collins prefer -ize spellings, which are more suitable
than -ise for international English.
(16) For American meanings or spellings that are sometimes different from
European, consult, for example, Webster's New Collegiate Dictionary of the
English Language, Merriam, Springfield,. MA, or the Concise Oxford
Dictionary.
(17) Roget's Thesaurus is invaluable in helping one to find a word. Several
editions are available. Differently arranged is Collins New World Thesau-
rus, by C. Laird (1979), Collins, London & Glasgow.
Every scientist should have a dictionary of science. A good one is the
Dictionary of Science & Technology (1988), Chambers, Edinburgh, and
Cambridge University Press.
There are other science dictionaries. To test one, look up items in your
own subject. If the entries are out of date, beware!
More a book to be read than consulted as a dictionary is Words of
Science, by Isaac Asimov (1974), Harrap, London .

Units & nomenclature

The source book of units is Le Systeme Internationale D' Unites (SJ),
(1985), Bureau International des Poids & Mesures, Sevres, France. The
book contains an English translation. More suitable for the research
scientist are the following book and leaflet.
(18) Quantitites, Units, & Symbols by the Symbols Committee (1975), The
Royal Society, London.
Quantities, Units & Symbol$ in Physical Chemistry, by K. H . Homan, is
an abbreviated list published by Blackwell Scientific Publications, Oxford.
A beautiful chart, Quantities & SJ Units, is printed in seven colours and
·published by Nederlands Normalisatie-instituut, Polakweg 5, Rijswijk ZH,
Netherlands.
If you have difficulty in understanding a unit as defined in one book
[many of us do], try reading about the unit in a different sort of book, for
example a science dictionary.
(19) BS3763 (1976) on SJ Units and BS4795 (1972) on All-Numeric Dates are
issued by The British Standards Institution, 2 Park St, London WlA 2BS.
 FURTHER READING 73

Rules for nomenclature of chemicals are given in Nomenclature of

Inorganic Chemistry, by IUPAC (International Union of Pure & Applied
Chemistry) (l989), Butterworths, London, and Nomenclature of Organic
Chemistry (1979), Pergamon, Oxford. Some of the information in the latter
is in Biochemical Nomenclature & Related Documents (1978), Biochemical
Society, London & Colchester, and the price is much lower.
For a quick check on spelling of chemical names you may consult a
(20) well-produced catalogue of chemicals. Such catalogues include Laboratory
Chemicals, BDH Chemicals Ltd, Poole, Dorset BH12 4NN.

Date of printing
The dates given for the books are the dates of issue. A date may be that of
a reprinting of an older edition.
Index
Suffixes a, b, c refer to top, middle and lowest thirds of a page, respectively.

abbreviation Sc, 16c, 23b, S2b comparative, comparatively 9c, 3Sb,

about 1Sa,23c S2b
acronyms 23b concentrations, chemical 25b
adjectives as nouns 8a , 18b concise Sa , 64a
aliquot 18b conclusion 2c, 3a, 6c, 30a, 37a, 67a
alternate, alternative 56a, SSb constantly 3Sa
American pronunciation 59b contractions 23b
American words S6a, 6lb, 62c control (as a noun) Sb
Americanism 5Sb control (as a verb) 50c
amplifier, speech 38c corn 56b
&20b correction
and, comma before 20c, 2la numbers42b
anderm t tb , 33a proof xiia
archaic names 17a, 25b, 3Sb script 47a, 65b, 67b
words&
Baer 55a could SOc
balloon xiia, 46a, 47b critical reading Sc, 26b
basically 33a, S6b curves 48c, 66c
beginning to write 2c
de Bono6b dangling participle lOb, 6lb
bred 57c dangling pronoun 12a, 14c
Butcher48a dangling verb lOc
dash 2lb
capita, per 16c data 62b
caption xvc, 23c dates 25c
case 13b Day 12c, 13b, 45c, (J)b
cf. 23c deafness, partial 34b
changeinlanguage18a,62c decimal S3b
chemicals, names of Sa, 22a, 73a demonstrate 15b, 5la
Clark lb, 67b diagrams 2Sa, 48c, 66c
clearly l 7b, see also plainly dictionary 48a, 52c, S3a, S6b, 61b, 72a
clicht 3a, 33b diet, feed a S7c
coefficient of variation, c.v. 5c, 24c different 13c
colon 19c, 20a digits lb, 16c, 24b, 4Ia
comma 13c, 1S4a, 20b dilution 4lc, 42a
discussion 2a, 3b, 6a, 19a, 23c
75
76 INDEX

Dixon, 8. 17a headline, running title 3c

Dixon, H . 8. F. 25c hedging Ile
dot-matrix printer 45b Herbert26c
drawings 48c Hers49b
due to 13c, 58c Hildebrand SSa
Holman60c
efficient 14a homonym 16a, l8a, 56c
em rule 21b hopefully 58c
emotion 26c, 31a Houp & Pearsall I la, 60b, 62a
employ 15a humility 27a, 61b
empty numbers 41a humour39a
empty words 3c, Ila, 19b, 32c, 33b , 62a hyphen 9c , 21b, 22a, 45c, 46a
en rule 21b
ensure 58a I 12b, 36a
episcope 37c illustrations 25a
error in vivo and in vitro 9c
in numbers 42b indentation 45a
in script 47a, 61b -ing word !Ob
in words lOa, SOa interesting 27a
every, one in 4lc interpreter 32a
export 56a Introduction 2a, 3a, 4b, 19a, 24a, 66c
eyestrain 46c invariably 35a
it 12a, 36a, 60c
fact l la, 14a, 26b -ization 18b, 55b, 62b
faucet 6lb
figure 16b, 48b Kenny39b
figure (opine) 61c key words 3b, 34a , 53b
final, finally 7a, 27b
fix 56c Ill 48a
footnote 44c L for litre 26a
foreign words !6c, 35b, SOb lateral thinking 6b
Fowler !Ob, 13b legend xvc, 44c, 49b
fractions 23a lettering 38b , 40b , 49b
full stop, period !9c, 23b localize 57a
future tense SSb logarithmic scale 66c
logic 6a, 67c
genitive 9a long adjectival phrases, see stacked
gerund lOb modifiers
goobledegook 6!a long words !Sa, 34c, 55c
grammar Sa, 53c, 54c loudspeakers 38c
grandiloquence 7b, 36c, 39c, Sia, 55c,
65a Mackay6c
graph lSb, 25a, 30a, 48c magnifier 37c, 49c
Maier6a
hackneyed phrases 3a, 33b margin 44c, 66b
Hartree 4c mask, for projector 30c, 38a, 40a
Harvard System 48b mathematical terms 15b, 23a
Hawkins67b method 2a, 3a, Sa, 29c
heading 23c, 24a, 65c µg45a
 INDEX 77

microphone 38c Partridge 13b

modifiers, stacked, see stacked Pauli 6c
modifiers percentage 26a, 43b
mo11 - 1 16b, 25b period, full stop 19c, 23b
molar26b permissiveness 53a
molecular weight 25c perspective 35a
Perttunen 7c
needless syllables 60c photocopy xva, 45b, 46b
needless words 3a, Jc, I la, l 9b, 32c, plainly lib, l7b
33a, 60a, 62a pomposity 7b, S5c, 60b
negative !Sc possessive case 9a
nervousness 3 la practicable 57b
Nicholson 60c practice, practise 58a
non-word 33a prefix 22b, 23a, 24b
normal 16a, 35b prior to 56a
Norman 17b projector 37b, 38a, 40a
Norris 37b promises 27 b
notebook la pronoun 12a, 14c, 36a, 60c
notes 30b, 3lc pronunciation 22b, 33c, 52c, 59b, 62b
noun adjective Ba, 9a, 18b, 22a, 54c, proof correction xiia
55a pulse rate 32a
NP xiic, 19c, 45a . punctuation l 9c, 67 b
number 16c, 18b, 42b , 47b , 52b
numbering 48b quantities, unscientific 4lb, 43a
numeral 16b, 23a question 4b, 6a, 24a
quite l 7b
oddity
dangling participle !Oa range instead of mean Sc
homonym 16a rapid reading 8b
not good sense 26b ratios, doubtful 4 lb
noun as adjective SSa reading a talk 30a, 39c
rabbit translated as dog 4c recent I lb
using and 20b referee 27b, 44a, 57b, 70c
using case l 3b references 2a, 4b, 7a, 44c, 45b, 48b, 67b
using diet 57c relative molecular mass 25c
using patient 52b repair 51a
of9a repetition I le, lSc, l9a, 24a, 26a
order, operational 5a reservoirs 2a, 30b
overhead projector 37b, 38a, 40a respectively 52a
owing to 13c, 58c results 2a, 5c, 29c, 62b
reverse meaning l 7b
page number 45b, 47b
.
review
paradox 4la article 4b
paragraph xiic, 19c, 4Sa referee 57b
paranym 17b revision and editing Be, 18c, 24a, 66a,
parenthesis 22c 67b
partial deafness 34b right here 62a
participle IOa rocks and stones 58a
parting remarks 7a Roland 4c
78 INDEX

running title, headline 3c thermistor 4Qa

time 25c
sanscrifxvib, 26a, 47c times more than 17c, 4lb
Schoenfeld I la Title 3b, 24a, 29a
semicolon 19c, 20b top sheet 41a
sentence 7c, 20a, 23a, 3lc, 39c trade names Sa
show l lb, 1Sb, Sta translation, incorrect 4c
SI units 2Sb, 72c transparencies 30c, 38a, 40a, 49b
significant digits 24b, 4 la transport(a ti on) S7b
skeleton, outline for paper 2a T. W . Aine xiva, 7c, 9b, l6b, 22b, 33c,
slides 30c, 37b, 39c S4b, 55c, S8b, 63a
so lib, S1a type face for typewriter or word
solidus, slash 22c, 23a processor 47c
spacing44c
speaking at meetings le, 28b, S3b units 23b, 24b, 2Sb , S3b, 72c
rate of 32b, 34b units, archaic 17a, 25b, 35b
verbal style 3lc units, named after people 26a
spelling 19b, S2c, S4b, 61b, see also
American words, Americanism variance Sc
stacked modifiers 9c, S4c, S6a, 60b varying, various 14b,26b
stage fright 3 la VDU46c
standard deviation, s.o. Sc, 24c, 2Sa , verbiage l 9b
42b very 17b, 3Sc
standard error of the mean, S.E. M. Sc, videotape 39b
18b,24b,2Sa,42b,67c visual aids 37b
starting to write 2c vocabulary7b, 26c, 6lb
statistics Sc, 29c, 42b, Slb, 61b vogue words 34c, S6b
Strunk Ile, 14b, 27a, 61a
Summary lb, 2a, 3b, 4a, 6a, 23c, 66c waUboard 30c, 34a, 37b, 38c
sure, surely 17b waste words 3c, 1 la, 32c, 33a
surprise 27a we 12b, 36b
synonym ISc which and that 14b
wolf words 35c
tables 24b, 37b, 44c, 48b, 66c Wolner26a
tape, self-adhesive 47b Woodford 2a , 9c, S4c, 60b
tautology I le, 32c, 62b word, choice of 13b, 26c, 34c, 3Sa, SOb ,
technical words 33c, 60a 58c
teleology 36b word division 4Sc
television 36b, 39b word processor 46b, 41c, 48b
tense 12c, SSb
that and which l 4b Yamane4c
therefore 1lb, 57a
Pri nted in the United States
82063L V00002B/ 1-2 19
II 111111111111
9 780521 429153
 ISBN 0- 521 - 42915-3
CAMBRIDGE
UNIVERSITY PRESS
w ww.cambridge.org